AI For Student Recruitment

AI For Student Recruitment — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

• 

  Cinema 4D

  Cinema 4D is a 3D software suite developed by the German company Maxon. == Overview == As of R21, only a single version of Cinema 4D is available. It replaces all previous variants, including BodyPaint 3D, and includes all features of the past 'Studio' variant. With R21, all binaries were unified. There is no technical difference between commercial, educational, or demo versions. The difference is now only in licensing. 2014 saw the release of Cinema 4D Lite, which came packaged with Adobe After Effects Creative Cloud 2014. "Lite" acts as an introductory version, with many features withheld. This is part of a partnership between the two companies, where a Maxon-produced plug-in, called Cineware, allows any variant to create a seamless workflow with After Effects. The "Lite" variant is dependent on After Effects CC, needing the latter application running to launch, and is only sold as a package component included with After Effects CC through Adobe. Initially, Cinema 4D was developed for Amiga computers in the early 1990s, and the first three versions of the program were available exclusively for that platform. With v4, however, Maxon began to develop the application for Windows and Macintosh computers as well, citing the wish to reach a wider audience and the growing instability of the Amiga market following Commodore's bankruptcy. It was also released for BeOS. On Linux, Cinema 4D is available as a commandline rendering version. == Modules and older variants == From R12 to R20, Cinema 4D was available in four variants. A core Cinema 4D 'Prime' application, a 'Broadcast' version with additional motion-graphics features, 'Visualize,' which adds functions for architectural design and 'Studio,' which includes all modules. From Release 8 until Release 11.5, Cinema 4D had a modular approach to the application, with the ability to expand upon the core application with various modules. This ended with Release 12, though the functionality of these modules remains in the different flavors of Cinema 4D (Prime, Broadcast, Visualize, Studio) The old modules were: Advanced Render (global illumination/HDRI, caustics, ambient occlusion and sky simulation) BodyPaint 3D (direct painting on UVW meshes; now included in the core. In essence Cinema 4D Core/Prime and the BodyPaint 3D products are identical. The only difference between the two is the splash screen that is shown at startup and the default user interface.) Dynamics (for simulating soft body and rigid body dynamics) Hair (simulates hair, fur, grass, etc.) MOCCA (character animation and cloth simulation) MoGraph (Motion Graphics procedural modelling and animation toolset) NET Render (to render animations over a TCP/IP network in render farms) PyroCluster (simulation of smoke and fire effects) Prime (the core application) Broadcast (adds MoGraph2) Visualize (adds Virtual Walkthrough, Advanced Render, Sky, Sketch and Toon, data exchange, camera matching) Studio (the complete package) == Version history == == Use in industry == A number of films and related works have been modeled and rendered in Cinema 4D, including: == Cinebench == Cinebench is a cross-platform test suite which tests a computer's hardware capabilities. It can be used as a test for Cinema 4D's 3D modeling, animation, motion graphic and rendering performance on multiple CPU cores. The program "target[s] a certain niche and [is] better suited for high-end desktop and workstation platforms". Cinebench is commonly used to demonstrate hardware capabilities at tech shows to show a CPU performance, especially by tech YouTubers and review sites.

  Read more →
• 

  Behavior selection algorithm

  In artificial intelligence, a behavior selection algorithm, or action selection algorithm, is an algorithm that selects appropriate behaviors or actions for one or more intelligent agents. In game artificial intelligence, it selects behaviors or actions for one or more non-player characters. Common behavior selection algorithms include: Finite-state machines Hierarchical finite-state machines Decision trees Behavior trees Hierarchical task networks Hierarchical control systems Utility systems Dialogue tree (for selecting what to say) == Related concepts == In application programming, run-time selection of the behavior of a specific method is referred to as the strategy design pattern.

  Read more →
• 

  Metadata

  Metadata (or metainformation) is data (or information) that defines and describes the characteristics of other data. It often helps to describe, explain, locate, or otherwise make data easier to retrieve, use, or manage. For example, the title, author, and publication date of a book are metadata about the book. But, while a data asset is finite, its metadata is infinite. As such, efforts to define, classify types, or structure metadata are expressed as examples in the context of its use. The term "metadata" has a history dating to the 1960s where it occurred in computer science and in popular culture. Different types of metadata serve different functions. For example, descriptive metadata for a document might include the author, creation date, file size and keywords. Metadata has various purposes. It can help users find relevant information and discover resources. It can also help organize electronic resources, provide digital identification, and archive and preserve resources. Metadata allows users to access resources by "allowing resources to be found by relevant criteria, identifying resources, bringing similar resources together, distinguishing dissimilar resources, and giving location information". Metadata of telecommunication activities including Internet traffic is very widely collected by various national governmental organizations. This data is used for the purposes of traffic analysis and can be used for mass surveillance. Unique metadata standards exist for different disciplines (e.g., museum collections, digital audio files, websites, etc.). Describing the contents and context of data or data files increases its usefulness. For example, a web page may include metadata specifying what software language the page is written in (e.g., HTML), what tools were used to create it, what subjects the page is about, and where to find more information about the subject. This metadata can automatically improve the reader's experience and make it easier for users to find the web page online. A CD may include metadata providing information about the musicians, singers, and songwriters whose work appears on the disc. In many countries, government organizations routinely store metadata about emails, telephone calls, web pages, video traffic, IP connections, and cell phone locations. == Types == There are many distinct types of metadata, including: Descriptive metadata – the descriptive information about a resource. It is used for discovery and identification. It includes elements such as title, abstract, author, and keywords. Structural metadata – metadata about containers of data and indicates how compound objects are put together, for example, how pages are ordered to form chapters. It describes the types, versions, relationships, and other characteristics of digital materials. Administrative metadata – the information to help manage a resource, like resource type, and permissions, and when and how it was created. Reference metadata – the information about the contents and quality of statistical data. Statistical metadata – also called process data, may describe processes that collect, process, or produce statistical data. Legal metadata – provides information about the creator, copyright holder, and public licensing, if provided. Metadata is not strictly bound to one of these categories, as it can describe a piece of data in many other ways. While the metadata application is manifold, covering a large variety of fields, there are specialized and well-accepted models to specify types of metadata. Bretherton & Singley (1994) distinguish between two distinct classes: structural/control metadata and guide metadata. Structural metadata describes the structure of database objects such as tables, columns, keys and indexes. Guide metadata helps humans find specific items and is usually expressed as a set of keywords in a natural language. According to Ralph Kimball, metadata can be divided into three categories: technical metadata (or internal metadata), business metadata (or external metadata), and process metadata. Dan Linstedt, creator of the data vault methodology, says business metadata "...provide[s] definition of the functionality, definition of the data, definition of the elements, and definition of how the data is used within business...business metadata includes business requirements, time-lines, business metrics, business process flows, and business terminology." Business metadata is important because it can greatly facilitate the usefulness of the data to business people. A simple example of business metadata is a glossary entry. Hover functionality in an application or web form can enable a glossary definition to be shown when cursor is on a field or term. Other examples of business metadata include annotation ability within applications. For example, a business user may be viewing a business intelligence (BI) report and notice a trend in the data. The user may have background knowledge as to why this trend occurs. Some business intelligence tools enable the user to create an annotation within the report that explains the trend. Such an annotation can enhance other users' understanding of the data. This example is especially powerful because it is created by a business user for the use of other business people. NISO distinguishes three types of metadata: descriptive, structural, and administrative. Descriptive metadata is typically used for discovery and identification, as information to search and locate an object, such as title, authors, subjects, keywords, and publisher. Structural metadata describes how the components of an object are organized. An example of structural metadata would be how pages are ordered to form chapters of a book. Finally, administrative metadata gives information to help manage the source. Administrative metadata refers to the technical information, such as file type, or when and how the file was created. Two sub-types of administrative metadata are rights management metadata and preservation metadata. Rights management metadata explains intellectual property rights, while preservation metadata contains information to preserve and save a resource. Statistical data repositories have their own requirements for metadata in order to describe not only the source and quality of the data but also what statistical processes were used to create the data, which is of particular importance to the statistical community in order to both validate and improve the process of statistical data production. An additional type of metadata beginning to be more developed is accessibility metadata. Accessibility metadata is not a new concept to libraries; however, advances in universal design have raised its profile. Projects like Cloud4All and GPII identified the lack of common terminologies and models to describe the needs and preferences of users and information that fits those needs as a major gap in providing universal access solutions. Those types of information are accessibility metadata. The Schema.org website has incorporated several accessibility properties based on IMS Global Access for All Information Model Data Element Specification. While the efforts to describe and standardize the varied accessibility needs of information seekers are beginning to become more robust, their adoption into established metadata schemas has not been as developed. For example, while Dublin Core (DC)'s "audience" and MARC 21's "reading level" could be used to identify resources suitable for users with dyslexia and DC's "format" could be used to identify resources available in braille, audio, or large print formats, there is more work to be done. == History == Metadata was traditionally used in the card catalogs of libraries until the 1980s when libraries converted their catalog data to digital databases. In the 2000s, as data and information were increasingly stored digitally, this digital data was described using metadata standards. An early description of "meta data" for computer systems was written by David Griffel and Stuart McIntosh at the MIT Center for International Studies in 1967: "In summary then, we have statements in an object language about subject descriptions of data and token codes for the data. We also have statements in a meta language describing the data relationships and transformations, and ought/is relations between norm and data." == Definition == Metadata means "data about data". Metadata is defined as the data providing information about one or more aspects of the data; it is used to summarize basic information about data that can make tracking and working with specific data easier. Some examples include: Means of creation of the data Source of the data Time and date of creation Creator or author of the data Location on a computer network where the data was created Standards used Data quality For example, a digital image may include metadata that describes the size of the image, its color depth, resolution,

  Read more →
• 

  Lai–Robbins lower bound

  The Lai–Robbins lower bound gives an asymptotic lower bound on the regret that any uniformly good algorithm must incur in the stochastic multi-armed bandit problem. The original result was proved by Tze Leung Lai and Herbert Robbins in 1985 for parametric exponential families. Later work extended the statement to more general classes of distributions. == Multi-armed bandit problem == The multi-armed bandit problem (MAB) is a sequential game in which the player must trade off exploration (to learn) and exploitation (to earn). The player chooses among K {\displaystyle K} actions (arms) with unknown distributions ν = ( ν 1 , … , ν K ) {\displaystyle \nu =(\nu _{1},\dots ,\nu _{K})} . The player is assumed to know a class of distributions D {\displaystyle {\mathcal {D}}} such that for every k {\displaystyle k} one has ν k ∈ D {\displaystyle \nu _{k}\in {\mathcal {D}}} (for example, D {\displaystyle {\mathcal {D}}} may be the family of Gaussian or Bernoulli distributions). At each round t = 1 , … , T {\displaystyle t=1,\dots ,T} the player selects (pulls) an arm a t {\displaystyle a_{t}} and observes a reward X t ∼ ν a t {\displaystyle X_{t}\sim \nu _{a_{t}}} . We denote N a ( t ) := ∑ s = 1 t 1 { a s = a } {\displaystyle N_{a}(t):=\sum _{s=1}^{t}\mathbf {1} _{\{a_{s}=a\}}} the number of times arm a {\displaystyle a} has been pulled in the first t {\displaystyle t} rounds, μ ( ν ) := ( μ 1 , … , μ K ) {\displaystyle \mu (\nu ):=(\mu _{1},\dots ,\mu _{K})} the vector of arm means, where μ k = E X ∼ ν k [ X ] {\displaystyle \mu _{k}=\mathbb {E} _{X\sim \nu _{k}}[X]} , μ ∗ := max a μ a {\displaystyle \mu ^{}:=\max _{a}\mu _{a}} the highest mean Δ a := μ ∗ − μ a ≥ 0 {\displaystyle \Delta _{a}:=\mu ^{}-\mu _{a}\geq 0} the gap of arm a {\displaystyle a} . An arm a {\displaystyle a} with μ a = μ ∗ {\displaystyle \mu _{a}=\mu ^{}} is called an optimal arm; otherwise it is a suboptimal arm. The goal is to minimize the regret at horizon T {\displaystyle T} , defined by R T := ∑ a = 1 K Δ a E [ N a ( T ) ] . {\displaystyle R_{T}:=\sum _{a=1}^{K}\Delta _{a}\,\mathbb {E} [N_{a}(T)].} Intuitively, the regret is the (expected) total loss compared to always playing an optimal arm: regret = ∑ a ( cost of playing a ) × ( times a is played ) . {\displaystyle {\text{regret}}=\sum _{a}\ ({\text{cost of playing }}a)\times ({\text{times }}a{\text{ is played}}).} An MAB algorithm is a (possibly randomized) policy that, at each round t {\displaystyle t} , choose an arm a_t by using the observations received from previous turns. === Intuitive example === Suppose a farmer must choose, each year, one of K {\displaystyle K} seed varieties to plant. Each variety k {\displaystyle k} has an unknown average yield μ k {\displaystyle \mu _{k}} . If the farmer knew the best variety (with mean μ ∗ {\displaystyle \mu ^{}} ) he would plant it every year; in reality he must try varieties to learn which is best. The cumulative regret after T {\displaystyle T} years measures the total expected loss in yield due to imperfect knowledge. Remarks The model above is the stochastic MAB; there also exist adversarial variants. One may consider a fixed-horizon setting (known T {\displaystyle T} ) or an anytime setting (unknown T {\displaystyle T} ). == Lai–Robbins lower bound == The theorem gives the right amount of time we should pull a suboptimal arm k {\displaystyle k} to distinguish whether we are in the instance with ν k {\displaystyle \nu _{k}} or with ν ~ k {\displaystyle {\tilde {\nu }}_{k}} where ν ~ k {\displaystyle {\tilde {\nu }}_{k}} is such that μ ~ k > μ ∗ {\displaystyle {\tilde {\mu }}_{k}>\mu ^{}} . Knowning a lower bound on the number of pull of every suboptimal arm gives a lower bound on the regret as only suboptimal arms contribute to the regret. Before stating the formal theorem we need to define what is a consistent algorithm. === Consistency (uniformly good algorithms) === Let D {\displaystyle {\mathcal {D}}} be a class of probability distributions and consider K {\displaystyle K} arms with reward distributions ν = ( ν 1 , … , ν K ) ∈ D K {\displaystyle \nu =(\nu _{1},\dots ,\nu _{K})\in {\mathcal {D}}^{K}} . An algorithm is said to be consistent (also called uniformly good) on D K {\displaystyle {\mathcal {D}}^{K}} if, for every instance ν ∈ D K {\displaystyle \nu \in {\mathcal {D}}^{K}} , the expected regret R T ( ν ) {\displaystyle R_{T}(\nu )} grows subpolynomially: ∀ α > 0 , R T ( ν ) = o ( T α ) as T → ∞ {\displaystyle \forall \alpha >0,\qquad R_{T}(\nu )=o(T^{\alpha })\quad {\text{as }}T\to \infty } This assumption excludes algorithms that perform well on some instances but incur linear regret on others. === Formal lower bound === For any suboptimal arm a {\displaystyle a} . For a distribution ν a ∈ D {\displaystyle \nu _{a}\in {\mathcal {D}}} and a threshold x {\displaystyle x} , define K inf ( ν a , x , D ) := inf { KL ⁡ ( ν a , ν ′ ) : ν ′ ∈ D , μ ′ > x } {\displaystyle {\mathcal {K}}_{\inf }(\nu _{a},x,{\mathcal {D}}):=\inf {\Bigl \{}\operatorname {KL} (\nu _{a},\nu '):\nu '\in {\mathcal {D}},\ \mu '>x{\Bigr \}}} where KL ⁡ ( ⋅ , ⋅ ) {\displaystyle \operatorname {KL} (\cdot ,\cdot )} denotes the Kullback-Leibler divergence. Then, for any algorithm consistent on D K {\displaystyle {\mathcal {D}}^{K}} and for every instance ν ∈ D K {\displaystyle \nu \in {\mathcal {D}}^{K}} , every suboptimal arm a {\displaystyle a} satisfies E ν [ N a ( T ) ] ≥ ln ⁡ T K inf ( ν a , μ ∗ , D ) + o ( ln ⁡ T ) {\displaystyle \mathbb {E} _{\nu }[N_{a}(T)]\geq {\frac {\ln T}{{\mathcal {K}}_{\inf }(\nu _{a},\mu ^{},{\mathcal {D}})}}+o(\ln T)} Consequently, the regret satisfies R T ( ν ) ≥ ( ∑ a : μ a < μ ∗ Δ a K inf ( ν a , μ ∗ , D ) ) ln ⁡ T + o ( ln ⁡ T ) {\displaystyle R_{T}(\nu )\geq \left(\sum _{a:\,\mu _{a}<\mu ^{}}{\frac {\Delta _{a}}{{\mathcal {K}}_{\inf }(\nu _{a},\mu ^{},{\mathcal {D}})}}\right)\ln T+o(\ln T)} The original 1985 paper established this result for exponential families; later work showed that the bound holds under much weaker assumptions on D {\displaystyle {\mathcal {D}}} . === Intuition === Consistency imposes that, for every ν {\displaystyle \nu } , the number of pulls of an optimal arm must be large. This means that μ ∗ {\displaystyle \mu ^{}} is estimated very accurately. The goal is to determine, for a suboptimal arm k {\displaystyle k} , how many samples are needed to be confident, with the appropriate level of confidence, that μ k < μ ∗ {\displaystyle \mu _{k}<\mu ^{}} . To do so, we use what is called the most confusing instance: an instance close to ν {\displaystyle \nu } such that arm k {\displaystyle k} is optimal. We define it as ν ~ {\displaystyle {\tilde {\nu }}} such that, for all a ≠ k {\displaystyle a\neq k} , ν ~ a = ν a {\displaystyle {\tilde {\nu }}_{a}=\nu _{a}} , and ν ~ k {\displaystyle {\tilde {\nu }}_{k}} is chosen so that μ ~ k > μ ∗ {\displaystyle {\tilde {\mu }}_{k}>\mu ^{}} . The objective is to determine how many samples of arm k {\displaystyle k} are required to distinguish whether we are in the instance with ν k {\displaystyle \nu _{k}} or with ν ~ k {\displaystyle {\tilde {\nu }}_{k}} in terms of KL {\displaystyle \operatorname {KL} } distance. == Algorithms achieving the Lai–Robbins lower bound == Several algorithms are known to achieve the Lai–Robbins asymptotic lower bound under specific assumptions on the reward distribution class D {\displaystyle {\mathcal {D}}} . The following list summarizes a non-exhaustive list of algorithms matching the lower bound. == Extension to other problems == === Structured bandit === A more complexe is structured bandit where we know that the mean of each arm is in a set with some restriction. In this case we can prove a smaller lower bound that use the knowledge of this set. === Best arm identification (BAI) === A similar result has been proved for best arm identification, which is the same game except that, instead of minimizing the regret, the goal is to identify the best arm with probability 1 − δ {\displaystyle 1-\delta } using as few rounds as possible. === Reinforcement Learning (RL) === Similar results have been proved for regret minimization in average-reward reinforcement learning. The order is also ln ⁡ T {\displaystyle \ln T} , with a constant that depends on the problem.

  Read more →
• 

  Weight initialization

  In deep learning, weight initialization or parameter initialization describes the initial step in creating a neural network. A neural network contains trainable parameters that are modified during training: weight initialization is the pre-training step of assigning initial values to these parameters. The choice of weight initialization method affects the speed of convergence, the scale of neural activation within the network, the scale of gradient signals during backpropagation, and the quality of the final model. Proper initialization is necessary for avoiding issues such as vanishing and exploding gradients and activation function saturation. Note that even though this article is titled "weight initialization", both weights and biases are used in a neural network as trainable parameters, so this article describes how both of these are initialized. Similarly, trainable parameters in convolutional neural networks (CNNs) are called kernels and biases, and this article also describes these. == Constant initialization == We discuss the main methods of initialization in the context of a multilayer perceptron (MLP). Specific strategies for initializing other network architectures are discussed in later sections. For an MLP, there are only two kinds of trainable parameters, called weights and biases. Each layer l {\displaystyle l} contains a weight matrix W ( l ) ∈ R n l − 1 × n l {\displaystyle W^{(l)}\in \mathbb {R} ^{n_{l-1}\times n_{l}}} and a bias vector b ( l ) ∈ R n l {\displaystyle b^{(l)}\in \mathbb {R} ^{n_{l}}} , where n l {\displaystyle n_{l}} is the number of neurons in that layer. A weight initialization method is an algorithm for setting the initial values for W ( l ) , b ( l ) {\displaystyle W^{(l)},b^{(l)}} for each layer l {\displaystyle l} . The simplest form is zero initialization: W ( l ) = 0 , b ( l ) = 0 {\displaystyle W^{(l)}=0,b^{(l)}=0} Zero initialization is usually used for initializing biases, but it is not used for initializing weights, as it leads to symmetry in the network, causing all neurons to learn the same features. In this page, we assume b = 0 {\displaystyle b=0} unless otherwise stated. Recurrent neural networks typically use activation functions with bounded range, such as sigmoid and tanh, since unbounded activation may cause exploding values. (Le, Jaitly, Hinton, 2015) suggested initializing weights in the recurrent parts of the network to identity and zero bias, similar to the idea of residual connections and LSTM with no forget gate. In most cases, the biases are initialized to zero, though some situations can use a nonzero initialization. For example, in multiplicative units, such as the forget gate of LSTM, the bias can be initialized to 1 to allow good gradient signal through the gate. For neurons with ReLU activation, one can initialize the bias to a small positive value like 0.1, so that the gradient is likely nonzero at initialization, avoiding the dying ReLU problem. == Random initialization == Random initialization means sampling the weights from a normal distribution or a uniform distribution, usually independently. === LeCun initialization === LeCun initialization, popularized in (LeCun et al., 1998), is designed to preserve the variance of neural activations during the forward pass. It samples each entry in W ( l ) {\displaystyle W^{(l)}} independently from a distribution with mean 0 and variance 1 / n l − 1 {\displaystyle 1/n_{l-1}} . For example, if the distribution is a continuous uniform distribution, then the distribution is U ( ± 3 / n l − 1 ) {\displaystyle {\mathcal {U}}(\pm {\sqrt {3/n_{l-1}}})} . === Glorot initialization === Glorot initialization (or Xavier initialization) was proposed by Xavier Glorot and Yoshua Bengio. It was designed as a compromise between two goals: to preserve activation variance during the forward pass and to preserve gradient variance during the backward pass. For uniform initialization, it samples each entry in W ( l ) {\displaystyle W^{(l)}} independently and identically from U ( ± 6 / ( n l + 1 + n l − 1 ) ) {\displaystyle {\mathcal {U}}(\pm {\sqrt {6/(n_{l+1}+n_{l-1})}})} . In the context, n l − 1 {\displaystyle n_{l-1}} is also called the "fan-in", and n l + 1 {\displaystyle n_{l+1}} the "fan-out". When the fan-in and fan-out are equal, then Glorot initialization is the same as LeCun initialization. === He initialization === As Glorot initialization performs poorly for ReLU activation, He initialization (or Kaiming initialization) was proposed by Kaiming He et al. for networks with ReLU activation. It samples each entry in W ( l ) {\displaystyle W^{(l)}} from N ( 0 , 2 / n l − 1 ) {\displaystyle {\mathcal {N}}(0,2/n_{l-1})} . === Orthogonal initialization === (Saxe et al. 2013) proposed orthogonal initialization: initializing weight matrices as uniformly random (according to the Haar measure) semi-orthogonal matrices, multiplied by a factor that depends on the activation function of the layer. It was designed so that if one initializes a deep linear network this way, then its training time until convergence is independent of depth. Sampling a uniformly random semi-orthogonal matrix can be done by initializing X {\displaystyle X} by IID sampling its entries from a standard normal distribution, then calculate ( X X ⊤ ) − 1 / 2 X {\displaystyle \left(XX^{\top }\right)^{-1/2}X} or its transpose, depending on whether X {\displaystyle X} is tall or wide. For CNN kernels with odd widths and heights, orthogonal initialization is done this way: initialize the central point by a semi-orthogonal matrix, and fill the other entries with zero. As an illustration, a kernel K {\displaystyle K} of shape 3 × 3 × c × c ′ {\displaystyle 3\times 3\times c\times c'} is initialized by filling K [ 2 , 2 , : , : ] {\displaystyle K[2,2,:,:]} with the entries of a random semi-orthogonal matrix of shape c × c ′ {\displaystyle c\times c'} , and the other entries with zero. (Balduzzi et al., 2017) used it with stride 1 and zero-padding. This is sometimes called the Orthogonal Delta initialization. Related to this approach, unitary initialization proposes to parameterize the weight matrices to be unitary matrices, with the result that at initialization they are random unitary matrices (and throughout training, they remain unitary). This is found to improve long-sequence modelling in LSTM. Orthogonal initialization has been generalized to layer-sequential unit-variance (LSUV) initialization. It is a data-dependent initialization method, and can be used in convolutional neural networks. It first initializes weights of each convolution or fully connected layer with orthonormal matrices. Then, proceeding from the first to the last layer, it runs a forward pass on a random minibatch, and divides the layer's weights by the standard deviation of its output, so that its output has variance approximately 1. === Fixup initialization === In 2015, the introduction of residual connections allowed very deep neural networks to be trained, much deeper than the ~20 layers of the previous state of the art (such as the VGG-19). Residual connections gave rise to their own weight initialization problems and strategies. These are sometimes called "normalization-free" methods, since using residual connection could stabilize the training of a deep neural network so much that normalizations become unnecessary. Fixup initialization is designed specifically for networks with residual connections and without batch normalization, as follows: Initialize the classification layer and the last layer of each residual branch to 0. Initialize every other layer using a standard method (such as He initialization), and scale only the weight layers inside residual branches by L − 1 2 m − 2 {\displaystyle L^{-{\frac {1}{2m-2}}}} . Add a scalar multiplier (initialized at 1) in every branch and a scalar bias (initialized at 0) before each convolution, linear, and element-wise activation layer. Similarly, T-Fixup initialization is designed for Transformers without layer normalization. === Others === Instead of initializing all weights with random values on the order of O ( 1 / n ) {\displaystyle O(1/{\sqrt {n}})} , sparse initialization initialized only a small subset of the weights with larger random values, and the other weights zero, so that the total variance is still on the order of O ( 1 ) {\displaystyle O(1)} . Random walk initialization was designed for MLP so that during backpropagation, the L2 norm of gradient at each layer performs an unbiased random walk as one moves from the last layer to the first. Looks linear initialization was designed to allow the neural network to behave like a deep linear network at initialization, since W R e L U ( x ) − W R e L U ( − x ) = W x {\displaystyle W\;\mathrm {ReLU} (x)-W\;\mathrm {ReLU} (-x)=Wx} . It initializes a matrix W {\displaystyle W} of shape R n 2 × m {\displaystyle \mathbb {R} ^{{\frac {n}{2}}\times m}} by any method, such as orthogonal initialization, t

  Read more →
• 

  Skyline operator

  The skyline operator is the subject of an optimization problem and computes the Pareto optimum on tuples with multiple dimensions. This operator is an extension to SQL proposed by Börzsönyi et al. to filter results from a database to keep only those objects that are not dominated by any other point on all dimensions. The name skyline comes from the view on Manhattan from the Hudson River, where those buildings can be seen that are not hidden by any other. A building is visible if it is not dominated by a building that is taller or closer to the river (two dimensions, distance to the river minimized, height maximized). Another application of the skyline operator involves selecting a hotel for a holiday. The user wants the hotel to be both cheap and close to the beach. However, hotels that are close to the beach may also be expensive. In this case, the skyline operator would only present those hotels that are not worse than any other hotel in both price and distance to the beach. == Formal specification == The skyline operator returns tuples that are not dominated by any other tuple. A tuple dominates another if it is at least as good in all dimensions and better in at least one dimension. Formally, we can think of each tuple as a vector p , q ∈ R n {\displaystyle p,q\in \mathbb {R} ^{n}} . p {\displaystyle p} dominates q {\displaystyle q} (written: p ≻ q {\displaystyle p\succ q} ) if p {\displaystyle p} is at least as good as q {\displaystyle q} in every dimension, and superior in at least one: p ≻ q ⇔ ∀ i ∈ [ n ] . p [ i ] ⪰ q [ i ] ∧ ∃ j ∈ [ n ] . p [ j ] ≻ q [ j ] . {\displaystyle p\succ q\Leftrightarrow \forall i\in [n].p[i]\succeq q[i]\wedge \exists j\in [n].p[j]\succ q[j].} Dominance ( p ≻ q {\displaystyle p\succ q} ) can be defined as any strict partial ordering, for example greater (with ≻:=> {\displaystyle \succ :=>} and ⪰:=≥ {\displaystyle \succeq :=\geq } ) or less (with ≻:=< {\displaystyle \succ :=<} and ⪰:=≤ {\displaystyle \succeq :=\leq } ). Assuming two dimensions and defining dominance in both dimensions as greater, we can compute the skyline in SQL-92 as follows: == Proposed syntax == As an extension to SQL, Börzsönyi et al. proposed the following syntax for the skyline operator: where d1, ... dm denote the dimensions of the skyline and MIN, MAX and DIFF specify whether the value in that dimension should be minimised, maximised or simply be different. Without an SQL extension, the SQL query requires an antijoin with not exists: == Implementation == The skyline operator can be implemented directly in SQL using current SQL constructs, but this has been shown to be very slow in disk-based database systems. Other algorithms have been proposed that make use of divide and conquer, indices, MapReduce and general-purpose computing on graphics cards. Skyline queries on data streams (i.e. continuous skyline queries) have been studied in the context of parallel query processing on multicores, owing to their wide diffusion in real-time decision making problems and data streaming analytics. Exasol features a native implementation.

  Read more →
• 

  Object storage

  Object storage (also known as object-based storage or blob storage) is a computer data storage approach that manages data as "blobs" or "objects", as opposed to other storage architectures like file systems, which manage data as a file hierarchy, and block storage, which manages data as blocks within sectors and tracks. Each object is typically associated with a variable amount of metadata, and a globally unique identifier. Object storage can be implemented at multiple levels, including the device level (object-storage device), the system level, and the interface level. In each case, object storage seeks to enable capabilities not addressed by other storage architectures, like interfaces that are directly programmable by the application, a namespace that can span multiple instances of physical hardware, and data-management functions like data replication and data distribution at object-level granularity. Object storage systems allow retention of massive amounts of unstructured data in which data is written once and read once (or many times). Object storage is used for purposes such as storing objects like videos and photos on Facebook, songs on Spotify, or files in online collaboration services, such as Dropbox. One of the limitations with object storage is that it is not intended for transactional data, as object storage was not designed to replace NAS file access and sharing; it does not support the locking and sharing mechanisms needed to maintain a single, accurately updated version of a file. == History == === Origins === Jim Starkey coined the term blob working at Digital Equipment Corporation to refer to opaque data entities. The terminology was adopted for Rdb/VMS. Blob is often humorously explained to be an abbreviation for binary large object. According to Starkey, this backronym arose when Terry McKiever, working in marketing at Apollo Computer felt that the term needed to be an abbreviation. McKiever began using the expansion basic large object. This was later eclipsed by the retroactive explanation of blobs as binary large objects. According to Starkey, "Blob don't stand for nothin'." Rejecting the acronym, he explained his motivation behind the coinage, saying, "A blob is the thing that ate Cincinnatti [sic], Cleveland, or whatever", referring to the 1958 science fiction film The Blob. In 1995, research led by Garth Gibson on Network-Attached Secure Disks first promoted the concept of splitting less common operations, like namespace manipulations, from common operations, like reads and writes, to optimize the performance and scale of both. In the same year, a Belgian company – FilePool – was established to build the basis for archiving functions. Object storage was proposed at Gibson's Carnegie Mellon University lab as a research project in 1996. Another key concept was abstracting the writes and reads of data to more flexible data containers (objects). Fine grained access control through object storage architecture was further described by one of the NASD team, Howard Gobioff, who later was one of the inventors of the Google File System. Other related work includes the Coda filesystem project at Carnegie Mellon, which started in 1987, and spawned the Lustre file system. There is also the OceanStore project at UC Berkeley, which started in 1999 and the Logistical Networking project at the University of Tennessee Knoxville, which started in 1998. In 1999, Gibson founded Panasas to commercialize the concepts developed by the NASD team. === Development === Seagate Technology played a central role in the development of object storage. According to the Storage Networking Industry Association (SNIA), "Object storage originated in the late 1990s: Seagate specifications from 1999 Introduced some of the first commands and how operating system effectively removed from consumption of the storage." A preliminary version of the "OBJECT BASED STORAGE DEVICES Command Set Proposal" dated 10/25/1999 was submitted by Seagate as edited by Seagate's Dave Anderson and was the product of work by the National Storage Industry Consortium (NSIC) including contributions by Carnegie Mellon University, Seagate, IBM, Quantum, and StorageTek. This paper was proposed to INCITS T-10 (International Committee for Information Technology Standards) with a goal to form a committee and design a specification based on the SCSI interface protocol. This defined objects as abstracted data, with unique identifiers and metadata, how objects related to file systems, along with many other innovative concepts. Anderson presented many of these ideas at the SNIA conference in October 1999. The presentation revealed an IP Agreement that had been signed in February 1997 between the original collaborators (with Seagate represented by Anderson and Chris Malakapalli) and covered the benefits of object storage, scalable computing, platform independence, and storage management. == Architecture == === Abstraction of storage === One of the design principles of object storage is to abstract some of the lower layers of storage away from the administrators and applications. Thus, data is exposed and managed as objects instead of blocks or (exclusively) files. Objects contain additional descriptive properties which can be used for better indexing or management. Administrators do not have to perform lower-level storage functions like constructing and managing logical volumes to utilize disk capacity or setting RAID levels to deal with disk failure. Object storage also allows the addressing and identification of individual objects by more than just file name and file path. Object storage adds a unique identifier within a bucket, or across the entire system, to support much larger namespaces and eliminate name collisions. === Inclusion of rich custom metadata within the object === Object storage explicitly separates file metadata from data to support additional capabilities. As opposed to fixed metadata in file systems (filename, creation date, type, etc.), object storage provides for full function, custom, object-level metadata in order to: Capture application-specific or user-specific information for better indexing purposes Support data-management policies (e.g. a policy to drive object movement from one storage tier to another) Centralize management of storage across many individual nodes and clusters Optimize metadata storage (e.g. encapsulated, database or key value storage) and caching/indexing (when authoritative metadata is encapsulated with the metadata inside the object) independently from the data storage (e.g. unstructured binary storage) Additionally, in some object-based file-system implementations: The file system clients only contact metadata servers once when the file is opened and then get content directly via object-storage servers (vs. block-based file systems which would require constant metadata access) Data objects can be configured on a per-file basis to allow adaptive stripe width, even across multiple object-storage servers, supporting optimizations in bandwidth and I/O Object-based storage devices (OSD) as well as some software implementations (e.g., DataCore Swarm) manage metadata and data at the storage device level: Instead of providing a block-oriented interface that reads and writes fixed sized blocks of data, data is organized into flexible-sized data containers, called objects Each object has both data (an uninterpreted sequence of bytes) and metadata (an extensible set of attributes describing the object); physically encapsulating both together benefits recoverability. The command interface includes commands to create and delete objects, write bytes and read bytes to and from individual objects, and to set and get attributes on objects Security mechanisms provide per-object and per-command access control === Programmatic data management === Object storage provides programmatic interfaces to allow applications to manipulate data. At the base level, this includes Create, read, update and delete (CRUD) functions for basic read, write and delete operations. Some object storage implementations go further, supporting additional functionality like object/file versioning, object replication, life-cycle management and movement of objects between different tiers and types of storage. Most API implementations are REST-based, allowing the use of many standard HTTP calls. == Implementation == === Cloud storage === The vast majority of cloud storage available in the market leverages an object-storage architecture. Some notable examples are Amazon S3, which debuted in March 2006, Microsoft Azure Blob Storage, IBM Cloud Object Storage, Rackspace Cloud Files (whose code was donated in 2010 to Openstack project and released as OpenStack Swift), and Google Cloud Storage released in May 2010. === Object-based file systems === Some distributed file systems use an object-based architecture, where file metadata is stored in metadata servers and file data is stored i

  Read more →
• 

  Information scientist

  The term information scientist developed in the latter part of the twentieth century by Wm. Hovey Smith to describe an individual, usually with a relevant subject degree (such as one in Information and Computer Science - CIS) or high level of subject knowledge, providing focused information to scientific and technical research staff in industry. It is a role quite distinct from and complementary to that of a librarian. Developments in end-user searching, together with some convergence between the roles of librarian and information scientist, have led to a diminution in its use in this context, and the term information officer or information professional (information specialist) are also now used. The term was, and is, also used for an individual carrying out research in information science. Brian C. Vickery mentions that the Institute of Information Scientists (IIS) was established in London during 1958 and lists the criteria put forward by this institute "Criteria for Information Science" (appendix 1) as well as his own "Areas of study in information science" (appendix 2). The IIS merged with the Library Association in 2002 to form the Chartered Institute of Library and Information Professionals (CILIP). == Notable Information Scientists == See also Award of Merit - Association for Information Science and Technology Marcia Bates David Blair (information technologist) Samuel C. Bradford Michael Buckland John M. Carroll Blaise Cronin Emilia Currás Brenda Dervin Eugene Garfield Paul B. Kantor Frederick Wilfrid Lancaster Calvin Mooers Tefko Saracevic Linda C. Smith Robert Saxton Taylor Brian Campbell Vickery Thomas D. Wilson == Additional reading == Ellis, David and Merete Haugan. (1997) "Modelling the information seeking patterns of engineers and research scientists in an industrial environment" (Journal of Documentation, Volume 53(4): pp. 384–403) Poole, Alex H. (2024). "'There's a big difference between going through life with the wind at your back, and going through life leaning into the wind': Feminism in Post-World War II Information Science". Proceedings of the Association for Information Science and Technology. 61: 300–313. doi:10.1002/pra2.1029. Vickery, Brian Campbell (1988) "Essays presented to B. C. Vickery" (Journal of Documentation, Volume 44, pp. 199–283). Vickery, B. & Vickery, A. (1987) Information Science in theory and practice (London: Bowker-Saur, pp. 361–369)

  Read more →
• 

  Compute (machine learning)

  In machine learning and deep learning, compute is the amount of computing power or computational resources required to train machine learning models and large language models. More broadly, compute is the computational power or resources necessary for a computer or computer program to function. == Definition == Compute is commonly defined as the amount of computing power or computational resources required to train machine learning and large language models. The term "compute" has also been more broadly applied to cloud computing, referencing processing power, memory, networking, storage, and other resources required for the computation of any program. Compute is measured in petaflop/s-days and is used to document AI training. A petaflop/s-day (pfs-day) consists of performing 1015 neural net operations per second for one day, or a total of about 1020 operations. The compute-time product serves as a mental convenience, similar to kilowatt-hour for energy. An amount of compute is meant to give an idea of the number of actual operations performed. == History == In a 2018 analysis titled "AI and compute", artificial intelligence company OpenAI introduced the concept of compute. OpenAI identified two eras of training AI systems in terms of compute-usage. From 1959 to 2012, compute roughly followed Moore’s law. Between 2012 and 2018, the amount of compute used in the largest AI training runs increased exponentially, growing by more than 300,000 times — roughly doubling every 3.4 months. By comparison, Moore’s Law doubled every two years over the same period. One of the largest models, released in 2020, used 600,000 times more computing power than the 2012 model. After 2020, compute growth began to slow down, with the compute needed for the largest AI models continuing to slow down in 2023. The notion of compute has become increasingly used from the mid-2020s onwards. == Compute growth and AI progress == Larger AI models trained on more data and using more computational resources, tend to perform better. This happens even if the algorithms themselves remain unchanged. As early as 2018, OpenAI noted the exponential increase in compute to be have a key role in AI progress. OpenAI considers three factors drive the advance of AI: algorithmic innovation, data, and the amount of compute available for training. AI models with more compute not only improve in the tasks they were trained on but can develop emergent abilities. Incremental improvements can lead to more abrupt leaps in capabilities. AI provider SpaceXAI said in 2026 that their AI progress is driven by compute and used it a key metric in the AI training of its supercomputer Colossus, the which contains 1 million GPUs. Anthropic has a contract of $1.25 billion per month with SpaceXAI to buy all the compute capacity at Colossus 1 data center. === Criticism and policy === Increasing, promoting or constraining progress in artificial intelligence has often be done via controlling the amount of compute. Policymarkers have enacted policies and provided support to make compute resources more accessible to domestic AI researchers. In a January 2022 report, the Center for Security and Emerging Technology (CSET) suggested to institutions that increasingly powerful and generalizable AI (AGI) will likely require other strategies than maximizing compute. Some AI researchers are also concerned that government might exclusively focus on scaling compute instead of other strategies. The CSET has reported on the various bottlenecks which could explain why deep learning needs for compute have slow down: training is expensive and training extremely large models generates traffic jams across many processors that are difficult to manage. there is a limited supply of AI chips (see AI chip memory shortage). CSET advances that the main resource is human capital, specifically talented researchers — according to a 2023 published survey of more than 400 AI researchers, academic and private sector workers. The survey found that AI researchers are not primarily or exclusively constrained by compute access. However, both academic and industry AI researchers equally report concerns that insufficient compute could prevent them from contributing meaningfully to AI research in the future. High compute users are more concerned about compute access. When asked about which resource provided by the government would be the most useful to them, some AI researchers select compute, other prefer grant funding. For this goal, CSET advised policymakers to ensure that even researchers with smaller budgets could effectively contribute to AI research. Other proposed strategies include using contemporary AI algorithms, managing modern AI infrastructure or focusing on interdisciplinary work between the AI field and other fields of computer science. A 2024 study on compute access found that academic-only AI research teams often have less compute intensive research topics, especially foundation models, compared to industry AI labs. As a consequence, academia is likely to play a smaller role in advancing such techniques. The researchers suggest nationally-sponsored computing infrastructure as well as open science initiatives to boost academic compute access. === Data === A 2022 study found that current large language models are significantly under-trained, a consequence of focusing on scaling language models whilst keeping the amount of training data constant. By training over 400 language models of various parameter and token size, they found that "for compute-optimal training", the model size and the number of training tokens should ideally be scaled equally: for every doubling of model size the number of training tokens should also be doubled.

  Read more →
• 

  Research data archiving

  Research data archiving is the long-term storage of scholarly research data, including the natural sciences, social sciences, and life sciences. The various academic journals have differing policies regarding how much of their data and methods researchers are required to store in a public archive, and what is actually archived varies widely between different disciplines. Similarly, the major grant-giving institutions have varying attitudes towards public archiving of data. In general, the tradition of science has been for publications to contain sufficient information to allow fellow researchers to replicate and therefore test the research. In recent years this approach has become increasingly strained as research in some areas depends on large datasets which cannot easily be replicated independently. Data archiving is more important in some fields than others. In a few fields, all of the data necessary to replicate the work is already available in the journal article. In drug development, a great deal of data is generated and must be archived so researchers can verify that the reports the drug companies publish accurately reflect the data. Often used interchangeably, Data preservation and data archiving are both about protecting data for the long term, but they serve different purposes. Data preservation focuses on preventing data from being lost, damaged, or destroyed by creating backups, storing data in secure locations, and ensuring it remains accessible when needed. Data archiving, on the other hand, involves moving data that is no longer actively used to a separate storage location for long-term keeping. Archived data is often combined and compressed, and while it can still be accessed, it is not intended for regular use or frequent updates. The requirement of data archiving is a recent development in the history of science. It was made possible by advances in information technology allowing large amounts of data to be stored and accessed from central locations. For example, the American Geophysical Union (AGU) adopted their first policy on data archiving in 1993, about three years after the beginning of the WWW. This policy mandates that datasets cited in AGU papers must be archived by a recognised data center; it permits the creation of "data papers"; and it establishes AGU's role in maintaining data archives. But it makes no requirements on paper authors to archive their data. Prior to organized data archiving, researchers wanting to evaluate or replicate a paper would have to request data and methods information from the author. The academic community expects authors to share supplemental data. This process was recognized as wasteful of time and energy and obtained mixed results. Information could become lost or corrupted over the years. In some cases, authors simply refuse to provide the information. The need for data archiving and due diligence is greatly increased when the research deals with health issues or public policy formation. == Selected policies by journals == === Biotropica === Biotropica requires, as a condition for publication, that the data supporting the results in the paper and metadata describing them must be archived in an appropriate public archive such as Dryad, Figshare, GenBank, TreeBASE, or NCBI. Authors may elect to make the data publicly available as soon as the article is published or, if the technology of the archive allows, embargo access to the data up to three years after article publication. A statement describing Data Availability will be included in the manuscript as described in the instructions to authors. Exceptions to the required archiving of data may be granted at the discretion of the Editor-in-Chief for studies that include sensitive information (e.g., the location of endangered species). Our Editorial explaining the motivation for this policy can be found here. A more comprehensive list of data repositories is available here. Promoting a culture of collaboration with researchers who collect and archive data: The data collected by tropical biologists are often long-term, complex, and expensive to collect. The Board of Editors of Biotropica strongly encourages authors who re-use data archives archived data sets to include as fully engaged collaborators the scientists who originally collected them. We feel this will greatly enhance the quality and impact of the resulting research by drawing on the data collector’s profound insights into the natural history of the study system, reducing the risk of errors in novel analyses, and stimulating the cross-disciplinary and cross-cultural collaboration and training for which the ATBC and Biotropica are widely recognized. NB: Biotropica is one of only two journals that pays the fees for authors depositing data at Dryad. === The American Naturalist === The American Naturalist requires authors to deposit the data associated with accepted papers in a public archive. For gene sequence data and phylogenetic trees, deposition in GenBank or TreeBASE, respectively, is required. There are many possible archives that may suit a particular data set, including the Dryad repository for ecological and evolutionary biology data. All accession numbers for GenBank, TreeBASE, and Dryad must be included in accepted manuscripts before they go to Production. If the data is deposited somewhere else, please provide a link. If the data is culled from published literature, please deposit the collated data in Dryad for the convenience of your readers. Any impediments to data sharing should be brought to the attention of the editors at the time of submission so that appropriate arrangements can be worked out. === Journal of Heredity === The primary data underlying the conclusions of an article are critical to the verifiability and transparency of the scientific enterprise, and should be preserved in usable form for decades in the future. For this reason, Journal of Heredity requires that newly reported nucleotide or amino acid sequences, and structural coordinates, be submitted to appropriate public databases (e.g., GenBank; the EMBL Nucleotide Sequence Database; DNA Database of Japan; the Protein Data Bank; and Swiss-Prot). Accession numbers must be included in the final version of the manuscript. For other forms of data (e.g., microsatellite genotypes, linkage maps, images), the Journal endorses the principles of the Joint Data Archiving Policy (JDAP) in encouraging all authors to archive primary datasets in an appropriate public archive, such as Dryad, TreeBASE, or the Knowledge Network for Biocomplexity. Authors are encouraged to make data publicly available at time of publication or, if the technology of the archive allows, opt to embargo access to the data for a period up to a year after publication. The American Genetic Association also recognizes the vast investment of individual researchers in generating and curating large datasets. Consequently, we recommend that this investment be respected in secondary analyses or meta-analyses in a gracious collaborative spirit. === Molecular Ecology === Molecular Ecology expects that data supporting the results in the paper should be archived in an appropriate public archive, such as GenBank, Gene Expression Omnibus, TreeBASE, Dryad, the Knowledge Network for Biocomplexity, your own institutional or funder repository, or as Supporting Information on the Molecular Ecology web site. Data are important products of the scientific enterprise, and they should be preserved and usable for decades in the future. Authors may elect to have the data publicly available at time of publication, or, if the technology of the archive allows, may opt to embargo access to the data for a period up to a year after publication. Exceptions may be granted at the discretion of the editor, especially for sensitive information such as human subject data or the location of endangered species. === Nature === Such material must be hosted on an accredited independent site (URL and accession numbers to be provided by the author), or sent to the Nature journal at submission, either uploaded via the journal's online submission service, or if the files are too large or in an unsuitable format for this purpose, on CD/DVD (five copies). Such material cannot solely be hosted on an author's personal or institutional web site. Nature requires the reviewer to determine if all of the supplementary data and methods have been archived. The policy advises reviewers to consider several questions, including: "Should the authors be asked to provide supplementary methods or data to accompany the paper online? (Such data might include source code for modelling studies, detailed experimental protocols or mathematical derivations.) === Science === Science supports the efforts of databases that aggregate published data for the use of the scientific community. Therefore, before publication, large data sets (including microarray data, protein or DNA sequences, and atomic c

  Read more →
• 

  Weak stability boundary

  Weak stability boundary (WSB), including low-energy transfer, is a concept introduced by Edward Belbruno in 1987. The concept explained how a spacecraft could change orbits using very little fuel. Weak stability boundary is defined for the three-body problem. This problem considers the motion of a particle P of negligible mass moving with respect to two larger bodies, P1, P2, modeled as point masses, where these bodies move in circular or elliptical orbits with respect to each other, and P2 is smaller than P1. The force between the three bodies is the classical Newtonian gravitational force. For example, P1 is the Earth, P2 is the Moon and P is a spacecraft; or P1 is the Sun, P2 is Jupiter and P is a comet, etc. This model is called the restricted three-body problem. The weak stability boundary defines a region about P2 where P is temporarily captured. This region is in position-velocity space. Capture means that the Kepler energy between P and P2 is negative. This is also called weak capture. == Background == This boundary was defined for the first time by Edward Belbruno of Princeton University in 1987. He described a Low-energy transfer which would allow a spacecraft to change orbits using very little fuel. It was for motion about Moon (P2) with P1 = Earth. It is defined algorithmically by monitoring cycling motion of P about the Moon and finding the region where cycling motion transitions between stable and unstable after one cycle. Stable motion means P can completely cycle about the Moon for one cycle relative to a reference section, starting in weak capture. P needs to return to the reference section with negative Kepler energy. Otherwise, the motion is called unstable, where P does not return to the reference section within one cycle or if it returns, it has non-negative Kepler energy. The set of all transition points about the Moon comprises the weak stability boundary, W. The motion of P is sensitive or chaotic as it moves about the Moon within W. A mathematical proof that the motion within W is chaotic was given in 2004. This is accomplished by showing that the set W about an arbitrary body P2 in the restricted three-body problem contains a hyperbolic invariant set of fractional dimension consisting of the infinitely many intersections Hyperbolic manifolds. The weak stability boundary was originally referred to as the fuzzy boundary. This term was used since the transition between capture and escape defined in the algorithm is not well defined and limited by the numerical accuracy. This defines a "fuzzy" location for the transition points. It is also due the inherent chaos in the motion of P near the transition points. It can be thought of as a fuzzy chaos region. As is described in an article in Discover magazine, the WSB can be roughly viewed as the fuzzy edge of a region, referred to as a gravity well, about a body (the Moon), where its force of gravity becomes small enough to be dominated by force of gravity of another body (the Earth) and the motion there is chaotic. A much more general algorithm defining W was given in 2007. It defines W relative to n-cycles, where n = 1,2,3,..., yielding boundaries of order n. This gives a much more complex region consisting of the union of all the weak stability boundaries of order n. This definition was explored further in 2010. The results suggested that W consists, in part, of the hyperbolic network of invariant manifolds associated to the Lyapunov orbits about the L1, L2 Lagrange points near P2. The explicit determination of the set W about P2 = Jupiter, where P1 is the Sun, is described in "Computation of Weak Stability Boundaries: Sun-Jupiter Case". It turns out that a weak stability region can also be defined relative to the larger mass point, P1. A proof of the existence of the weak stability boundary about P1 was given in 2012, but a different definition is used. The chaos of the motion is analytically proven in "Geometry of Weak Stability Boundaries". The boundary is studied in "Applicability and Dynamical Characterization of the Associated Sets of the Algorithmic Weak Stability Boundary in the Lunar Sphere of Influence". == Applications == There are a number of important applications for the weak stability boundary (WSB). Since the WSB defines a region of temporary capture, it can be used, for example, to find transfer trajectories from the Earth to the Moon that arrive at the Moon within the WSB region in weak capture, which is called ballistic capture for a spacecraft. No fuel is required for capture in this case. This was numerically demonstrated in 1987. This is the first reference for ballistic capture for spacecraft and definition of the weak stability boundary. The boundary was operationally demonstrated to exist in 1991 when it was used to find a ballistic capture transfer to the Moon for Japan's Hiten spacecraft. Other missions have used the same transfer type as Hiten, including Grail, Capstone, Danuri, Hakuto-R Mission 1 and SLIM. The WSB for Mars is studied in "Earth-Mars Transfers with Ballistic Capture" and ballistic capture transfers to Mars are computed. The BepiColombo mission of ESA should achieve ballistic capture at the WSB of Mercury in November 2026. The WSB region can be used in the field of Astrophysics. It can be defined for stars within open star clusters. This is done in "Chaotic Exchange of Solid Material Between Planetary Systems: Implications for the Lithopanspermia Hypothesis" to analyze the capture of solid material that may have arrived on the Earth early in the age of the Solar System to study the validity of the lithopanspermia hypothesis. Numerical explorations of trajectories for P starting in the WSB region about P2 show that after the particle P escapes P2 at the end of weak capture, it moves about the primary body, P1, in a near resonant orbit, in resonance with P2 about P1. This property was used to study comets that move in orbits about the Sun in orbital resonance with Jupiter, which change resonance orbits by becoming weakly captured by Jupiter. An example of such a comet is 39P/Oterma. This property of change of resonance of orbits about P1 when P is weakly captured by the WSB of P2 has an interesting application to the field of quantum mechanics to the motion of an electron about the proton in a hydrogen atom. The transition motion of an electron about the proton between different energy states described by the Schrödinger equation is shown to be equivalent to the change of resonance of P about P1 via weak capture by P2 for a family of transitioning resonance orbits. This gives a classical model using chaotic dynamics with Newtonian gravity for the motion of an electron.

  Read more →
• 

  Australian Geoscience Data Cube

  The Australian Geoscience Data Cube (AGDC) is an approach to storing, processing and analyzing large collections of Earth observation data. The technology is designed to meet challenges of national interest by being agile and flexible with vast amounts of layered grid data. The AGDC reduces processing time of traditional image analysis by calibrating, pre-computing known extents, pixel alignment and storing metadata in a cell lattice structure. The temporal-pixel aligned data can often be analysed faster across space and time dimensions than previous scene based techniques. This allows the AGDC to be flexible in tackling future challenges and improve analysis times on every-increasing data repositories of earth observation. The AGDC has also been used internationally to allow countries to maintain ecologically sustainable programs and reduce the difficulty curve of utilizing Remote Sensing data. == Background == The AGDC was originally conceived by Geoscience Australia but is now maintained in a partnership between Geoscience Australia, Commonwealth Scientific and Industrial Research Organisation (CSIRO) and National Computational Infrastructure National Facility (Australia) (NCI). This is made possible by the funding from the partnership and a number of organisations such as National Collaborative Research Infrastructure Strategy (NCRIS). == Analysis ready data, ingestion and indexing == The data processed in the cube is made analysis ready before being ingested and indexed into the AGDC. Analysis ready data is pre-processed data that has applied corrections for instrument calibration (gains and offsets), geolocation (spatial alignment) and radiometry (solar illumination, incidence angle, topography, atmospheric interference). The ingestion process manages the translation of datasets into the storage units while maintaining a database index. The data within the storage and index can be accessed via API calls often compiled within code such as Python (programming language). Example: s2a_l1c = dc.load(product='s2a_level1c_granule',x=(147.36, 147.41), y=(-35.1, -35.15), measurements=['04','03','02'], output_crs='EPSG:4326', resolution=(-0.00025,0.00025)) === Datasets currently stored === Geoscience Australia Landsat Surface Reflectance (1987 to present) Landsat Pixel Quality Landsat Fractional Cover Landsat NDVI === Datasets that have been piloted === USGS Landsat Surface Reflectance SRTM DEM Himawari 8 MODIS Sentinel-2 L1C / S2A Australian Gridded Climate Data == Open source == The AGDC code base is situated in GitHub as an open repository. The core code base moved to the Open Data Cube in early 2017 as part of an international collaboration. Whilst the code base is the Open Data Cube, individual cubes exist as their own right such as the AGDC on the National Computational Infrastructure National Facility (Australia) (NCI) using the High-Performance Computing Cluster HPCC. The core code can be installed on personal computers or public computers (using git) and has many unit tests. Documentation for the code base exists on Read the Docs. == Challenges of the AGDC == The AGDC is designed to meet nationally significant challenges such as the following. Sustainability Environment Water resource management Disaster assist Policy development Community planning Forest preservation Carbon measurement == International awards == The AGDC won the 2016 Content Platform of the Year award from Geospatial World Forum.

  Read more →
• 

  Oculus Medium

  Oculus Medium is a digital sculpting software that works with virtual reality headsets and 6DoF motion controllers. It is used to create and paint digital sculptures. Medium works only on Oculus Rift. It was released on December 5, 2016, following with a major update in 2018 introducing new features and a revamped UI. On December 9, 2019, Oculus Medium was acquired by Adobe and re-named to "Medium by Adobe".

  Read more →
• 

  Subject indexing

  Subject indexing is the act of describing or classifying a document by index terms, keywords, or other symbols in order to indicate what different documents are about, to summarize their contents or to increase findability. In other words, the objective is to identify and describe the subject of documents. Indexes are constructed, separately, on three distinct levels: terms in a document, such as a book; objects in a collection, such as a library; and documents (such as books and articles) within a field of knowledge. Subject indexing is used in information retrieval especially to create bibliographic indexes to retrieve documents on a particular subject. Examples of academic indexing services are Zentralblatt MATH, Chemical Abstracts, and PubMed. The index terms were mostly assigned by experts but author keywords are also common. The process of indexing begins with any analysis of the subject of the document. The indexer must then identify terms that appropriately identify the subject, either by extracting words directly from the document or assigning words from a controlled vocabulary. The terms in the index are then presented in a systematic order. Indexers must decide how many terms to include and how specific the terms should be. Together this gives a depth of indexing. == Subject analysis == The first step in indexing is to decide on the subject matter of the document. In manual indexing, the indexer would consider the subject matter in terms of answer to a set of questions such as "Does the document deal with a specific product, condition or phenomenon?". As the analysis is influenced by the knowledge and experience of the indexer, it follows that two indexers may analyze the content differently and so come up with different index terms. This will impact on the success of retrieval. === Automatic vs. manual subject analysis === Automatic indexing follows set processes of analyzing frequencies of word patterns and comparing results to other documents in order to assign to subject categories. This requires no understanding of the material being indexed. This leads to more uniform indexing but at the expense of the true meaning being interpreted. A computer program will not understand the meaning of statements and may therefore fail to assign some relevant terms or assign incorrectly. Human indexers focus their attention on certain parts of the document such as the title, abstract, summary and conclusions, as analyzing the full text in depth is costly and time-consuming. An automated system takes away the time limit and allows the entire document to be analyzed, but also has the option to be directed to particular parts of the document. == Term selection == The second stage of indexing involves the translation of the subject analysis into a set of index terms. This can involve extracting from the document or assigning from a controlled vocabulary. With the ability to conduct a full text search widely available, many people have come to rely on their own expertise in conducting information searches and full text search has become very popular. Subject indexing and its experts, professional indexers, catalogers, and librarians, remains crucial to information organization and retrieval. These experts understand controlled vocabularies and are able to find information that cannot be located by full text search. The cost of expert analysis to create subject indexing is not easily compared to the cost of hardware, software and labor to manufacture a comparable set of full-text, fully searchable materials. With new web applications that allow every user to annotate documents, social tagging has gained popularity especially in the Web. One application of indexing, the book index, remains relatively unchanged despite the information revolution. === Extraction/Derived indexing === Extraction indexing involves taking words directly from the document. It uses natural language and lends itself well to automated techniques where word frequencies are calculated and those with a frequency over a pre-determined threshold are used as index terms. A stop-list containing common words (such as "the", "and") would be referred to and such stop words would be excluded as index terms. Automated extraction indexing may lead to loss of meaning of terms by indexing single words as opposed to phrases. Although it is possible to extract commonly occurring phrases, it becomes more difficult if key concepts are inconsistently worded in phrases. Automated extraction indexing also has the problem that, even with use of a stop-list to remove common words, some frequent words may not be useful for allowing discrimination between documents. For example, the term glucose is likely to occur frequently in any document related to diabetes. Therefore, use of this term would likely return most or all the documents in the database. Post-coordinated indexing where terms are combined at the time of searching would reduce this effect but the onus would be on the searcher to link appropriate terms as opposed to the information professional. In addition terms that occur infrequently may be highly significant for example a new drug may be mentioned infrequently but the novelty of the subject makes any reference significant. One method for allowing rarer terms to be included and common words to be excluded by automated techniques would be a relative frequency approach where frequency of a word in a document is compared to frequency in the database as a whole. Therefore, a term that occurs more often in a document than might be expected based on the rest of the database could then be used as an index term, and terms that occur equally frequently throughout will be excluded. Another problem with automated extraction is that it does not recognize when a concept is discussed but is not identified in the text by an indexable keyword. Since this process is based on simple string matching and involves no intellectual analysis, the resulting product is more appropriately known as a concordance than an index. === Assignment indexing === An alternative is assignment indexing where index terms are taken from a controlled vocabulary. This has the advantage of controlling for synonyms as the preferred term is indexed and synonyms or related terms direct the user to the preferred term. This means the user can find articles regardless of the specific term used by the author and saves the user from having to know and check all possible synonyms. It also removes any confusion caused by homographs by inclusion of a qualifying term. A third advantage is that it allows the linking of related terms whether they are linked by hierarchy or association, e.g. an index entry for an oral medication may list other oral medications as related terms on the same level of the hierarchy but would also link to broader terms such as treatment. Assignment indexing is used in manual indexing to improve inter-indexer consistency as different indexers will have a controlled set of terms to choose from. Controlled vocabularies do not completely remove inconsistencies as two indexers may still interpret the subject differently. == Index presentation == The final phase of indexing is to present the entries in a systematic order. This may involve linking entries. In a pre-coordinated index the indexer determines the order in which terms are linked in an entry by considering how a user may formulate their search. In a post-coordinated index, the entries are presented singly and the user can link the entries through searches, most commonly carried out by computer software. Post-coordination results in a loss of precision in comparison to pre-coordination. == Depth of indexing == Indexers must make decisions about what entries should be included and how many entries an index should incorporate. The depth of indexing describes the thoroughness of the indexing process with reference to exhaustivity and specificity. === Exhaustivity === An exhaustive index is one which lists all possible index terms. Greater exhaustivity gives a higher recall, or more likelihood of all the relevant articles being retrieved, however, this occurs at the expense of precision. This means that the user may retrieve a larger number of irrelevant documents or documents which only deal with the subject in little depth. In a manual system a greater level of exhaustivity brings with it a greater cost as more man-hours are required. The additional time taken in an automated system would be much less significant. At the other end of the scale, in a selective index only the most important aspects are covered. Recall is reduced in a selective index as if an indexer does not include enough terms, a highly relevant article may be overlooked. Therefore, indexers should strive for a balance and consider what the document may be used. They may also have to consider the implications of time and expense. === Specificity === The specificity describes how closel

  Read more →
• 

  Automated journalism

  Automated journalism, also known as algorithmic journalism or robot journalism, is a term that attempts to describe modern technological processes that are now in use in the journalistic profession, such as news articles and videos generated by computer programs. There are four main fields of application for automated journalism, namely automated content production, data mining, news dissemination and content optimization. Through generative artificial intelligence, stories are produced automatically by computers rather than human reporters. In the 2020s, generative pre-trained transformers have enabled the generation of articles, simply by providing prompts. Automated journalism is sometimes seen as an opportunity to free journalists from routine reporting, providing them with more time for complex tasks. It also allows efficiency and cost-cutting, alleviating some financial burden that many news organizations face. However, automated journalism is also perceived as a threat to the authorship and quality of news and a threat to the livelihoods of human journalists. == History == Historically, the process involved an algorithm that scanned large amounts of provided data, selected from an assortment of pre-programmed article structures, ordered key points, and inserted details such as names, places, amounts, rankings, statistics, and other figures. These programs interpret, organize, and present data in human-readable ways. The output can also be customized to fit a certain voice, tone, or style. Early implementations were mainly used for stories based on statistics and numerical figures. Common topics include sports recaps, weather, financial reports, real estate analysis, and earnings reviews. Data science and AI companies such as Automated Insights, Narrative Science, United Robots and Monok develop and provide these algorithms to news outlets. In 2016, early adopters included news providers such as the Associated Press, Forbes, ProPublica, and the Los Angeles Times. StatSheet, an online platform covering college basketball, runs entirely on an automated program. In 2006, Thomson Reuters announced their switch to automation to generate financial news stories on its online news platform. Reuters used a tool called Tracer. An algorithm called Quakebot published a story about a 2014 California earthquake on The Los Angeles Times website within three minutes after the shaking had stopped. The Associated Press began using automation to cover 10,000 minor baseball leagues games annually, using a program from Automated Insights and statistics from MLB Advanced Media. Outside of sports, the Associated Press also uses automation to produce stories on corporate earnings. Since 2014, Associated Press has been publishing quarterly financial stories with help from Automated Insights. In May 2020, Microsoft announced that a number of its MSN contract journalists would be replaced by robot journalism. On 8 September 2020, The Guardian published an article entirely written by the neural network GPT-3, although the published fragments were manually picked by a human editor. Agentic Tribune produces all of its news articles automatically using AI. News broadcasters in Kuwait, Greece, South Korea, India, China and Taiwan have presented news with anchors based on generative AI models, prompting concerns about job losses for human anchors and audience trust in news that has historically been influenced by parasocial relationships with broadcasters, content creators or social media influencers. Algorithmically generated anchors have also been used by allies of ISIS for their broadcasts. In 2023, Google reportedly pitched a tool to news outlets that claimed to "produce news stories" based on input data provided, such as "details of current events". Some news company executives who viewed the pitch described it as "[taking] for granted the effort that went into producing accurate and artful news stories." In February 2024, Google launched a program to pay small publishers to write three articles per day using a beta generative AI model. The program does not require the knowledge or consent of the websites that the publishers are using as sources, nor does it require the published articles to be labeled as being created or assisted by these models. Meta AI, a chatbot based on Llama 3 which summarizes news stories, was noted by The Washington Post to copy sentences from those stories without direct attribution and to potentially further decrease the traffic of online news outlets. == Benefits == === Speed === Robot reporters are built to produce large quantities of information at quicker speeds. The Associated Press announced that their use of automation has increased the volume of earnings reports from customers by more than ten times. With software from Automated Insights and data from other companies, they can produce 150 to 300-word articles in the same time it takes journalists to crunch numbers and prepare information. By automating routine stories and tasks, journalists are promised more time for complex jobs such as investigative reporting and in-depth analysis of events. Francesco Marconi of the Associated Press stated that, through automation, the news agency freed up 20 percent of reporters’ time to focus on higher-impact projects. This has also been stated by a spokesperson at Gannett, who stated "By leveraging AI, we are able to expand coverage and enable our journalists to focus on more in-depth reporting." GBH reports that AI tools help increase the reach of news publishers. Mike Carragi, a product manager at Patch, stated that they were able to increase their reach from 1200 communities to 7000 communities in just a few months without the need for new employees solely through the adoption of generative AI. In fact, many communities are served solely by AI generated content, which creates summaries of existing information within the community. === Cost === Automated journalism is cheaper because more content can be produced within less time. It also lowers labour costs for news organizations. Reduced human input means less expenses on wages or salaries, paid leaves, vacations, and employment insurance. Automation serves as a cost-cutting tool for news outlets struggling with tight budgets but still wish to maintain the scope and quality of their coverage. == Concerns == === Authorship === In an automated story, there is often confusion about who should be credited as the author. Several participants of a study on algorithmic authorship attributed the credit to the programmer; others perceived the news organization as the author, emphasizing the collaborative nature of the work. There is also no way for the reader to verify whether an article was written by a robot or human, which raises issues of transparency although such issues also arise with respect to authorship attribution between human authors too. === Credibility and quality === Concerns about the perceived credibility of automated news is similar to concerns about the perceived credibility of news in general. Critics doubt if algorithms are "fair and accurate, free from subjectivity, error, or attempted influence." Again, these issues about fairness, accuracy, subjectivity, error, and attempts at influence or propaganda has also been present in articles written by humans over thousands of years. A common criticism is that machines do not replace human capabilities such as creativity, humour, and critical-thinking. However, as the technology evolves, the aim is to mimic human characteristics. When the UK's Guardian newspaper used an AI to write an entire article in September 2020, commentators pointed out that the AI still relied on human editorial content. Austin Tanney, the head of AI at Kainos said: "The Guardian got three or four different articles and spliced them together. They also gave it the opening paragraph. It doesn’t belittle what it is. It was written by AI, but there was human editorial on that." The largest single study of readers' evaluations of news articles produced with and without the help of automation exposed 3,135 online news consumers to 24 articles. It found articles that had been automated were significantly less comprehensible, in part because they were considered to contain too many numbers. However, the automated articles were evaluated equally on other criteria including tone, narrative flow, and narrative structure. Beyond human evaluation, there are now numerous algorithmic methods to identify machine written articles although some articles may still contain errors that are obvious for a human to identify, they can at times score better with these automatic identifiers than human-written articles. A 2017 Nieman Reports article by Nicola Bruno discusses whether or not machines will replace journalists and addresses concerns around the concept of automated journalism practices. Ultimately, Bruno came to the conclusion that AI would assist journalist

  Read more →