AI Detector Just Done Free

AI Detector Just Done Free — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

ActivityPub

ActivityPub is a protocol and open standard for decentralized social networking. It provides a client-to-server (C2S) API for creating and modifying content, as well as a federated server-to-server (S2S) protocol for delivering notifications and content to other servers. ActivityPub is the defining standard of the Fediverse, a decentralised social network of various social interaction models, and content types, which consists of independently managed instances of software such as Mastodon, Pixelfed and PeerTube, among others. ActivityPub is considered to be an update to the ActivityPump protocol used in pump.io, and the official W3C repository for ActivityPub is identified as a fork of ActivityPump. The creation of a new standard for decentralized social networking was prompted by the complexity of OStatus, the most commonly used protocol at the time. OStatus was built using a multitude of technologies (such as Atom, Salmon, WebSub and WebFinger), a product of the infrastructure used in GNU social (the originator and largest user of the OStatus protocol), which made it difficult to implement the protocol into new software. OStatus was also only designed to work with microblogging services, with little flexibility to the types of data that it could hold. The standard was first published by the World Wide Web Consortium (W3C) as a W3C Recommendation in January 2018 by the Social Web Working Group (SocialWG), a working group chartered to build the protocols and vocabularies needed to create a standard for social functionality. Shortly after, further development was moved to the Social Web Community Group (SocialCG), the successor to the SocialWG. == Design == ActivityPub uses the ActivityStreams 2.0 format for building its content, which itself uses JSON-LD. The three main data types used in ActivityPub are Objects, Activities and Actors. Objects are the most common data type, and can be images, videos, or more abstract items such as locations or events. Activities are actions that create and modify objects, for example a Create activity creates an object. Actors are representative of an individual, a group, an application or a service, and are the owners of objects. Every actor type contains an inbox and outbox stream, which sends and receives activities for a user. In order to publish data (for example liking an article), a user creates an activity that declares that they liked an Article object and publishes it to their outbox, where it is then delivered by the ActivityPub server via a POST request to the inboxes listed in the activity's to, bto, cc and bcc fields. The receiving servers then account for the newly received activity and update the article by adding the like action to it. === Example data === An example actor object that represents a user account: An example activity that likes an article object: An example article object: == Project status == The SocialCG previously organized a yearly free conference called ActivityPub Conf about the future of ActivityPub. Triages are held regularly to review issues pertaining to the ActivityPub and ActivityStreams 2.0 specifications as part of the SocialCG. In 2023, Germany's Sovereign Tech Fund donated €152,000 to socialweb.coop with the goal of building a new suite for testing various ActivityPub implementations and their compliance with the specification. === Adoption === The initial wave of adoption for ActivityPub (circa 2016–2018) came from software that was already using OStatus as their federation protocol, such as Mastodon, GNU social and Pleroma. Following the acquisition of Twitter by Elon Musk in 2022, many groups of users that were critical of the acquisition migrated to Mastodon, bringing new attention to the ActivityPub protocol with it. Various major social media platforms and corporations have since pledged to implement ActivityPub support, including Tumblr, Flipboard and Meta Platforms' Threads. Threads introduced crossposting to ActivityPub in 2024 for users outside of the European Economic Area, however full 2-way compatibility remains incomplete as of 2025. == Criticism == === Accidental denial-of-service attacks === Poorly optimized ActivityPub implementations can cause unintentional distributed denial-of-service (DDOS) attacks on other websites and servers, due to the decentralized nature of the network. An example would be Mastodon's implementation of OpenGraph link previews, wherein every instance that receives a post that contains a link with OpenGraph metadata will download the associated data, such as a thumbnail, in a very short timeframe, which can slow down or crash servers as a result of the sudden burst of requests. === Account migration === ActivityPub has been criticized for not natively supporting moving accounts from one server to another, forcing implementations to build their own solutions. While there has been work on building a standardized system for migrating accounts using the Move activity via the Fediverse Enhancement Proposal organization, the current proposal only allows for basic follower migration, with all other data remaining linked to the original account. === Missing content and data === ActivityPub implementations have been criticized for missing replies and parts of reply threads from remote posts, and presenting outdated statistics (e.g. likes and reposts) about remote posts. However, this isn't a problem with the ActivityPub protocol itself, but with implementations not refreshing their content for updated data when needed. == Software using ActivityPub == === Future implementations === Flarum, an internet forum software Forgejo, a Git forge and development platform === Uncertain future implementations === GitLab, a Git forge and development platform which had previously had an open issue discussing the topic, but was later closed due to the development team moving focus to other areas. Tumblr, a microblogging platform. Despite previous statements from Automattic CEO Matt Mullenweg, ActivityPub integration has been delayed indefinitely. The integration would have been implemented with its WordPress migration, as the first-party plugin for interoperability would have been used for federation. Flickr, an image and video hosting site.
Read more →
Leo Breiman

Leo Breiman (January 27, 1928 – July 5, 2005) was an American statistician at the University of California, Berkeley and a member of the United States National Academy of Sciences. Breiman's work helped to bridge the gap between statistics and computer science, particularly in the field of machine learning. His most important contributions were his work on classification and regression trees and ensembles of trees fit to bootstrap samples. Bootstrap aggregation was given the name bagging by Breiman. Another of Breiman's ensemble approaches is the random forest.
Read more →
The Best Free AI Content Generator for Beginners

Looking for the best AI content generator? An AI content generator is software that uses machine learning to help you get more done — it can save you hours every week by automating repetitive work. Most options offer a generous free tier, with paid plans unlocking higher limits, faster processing, and team features. Whether you are a beginner or a pro, the right AI content generator slots into your workflow and pays for itself fast. This guide breaks down the top picks, their pros and cons, and who each one is best for.
Read more →
AI Clip Makers: Free vs Paid (2026)

Shopping for the best AI clip maker? An AI clip maker is software that uses machine learning to help you get more done — it keeps getting smarter as the underlying models improve. Pricing, accuracy, and the size of the model behind the tool are the three factors that most affect daily usefulness. Whether you are a beginner or a pro, the right AI clip maker slots into your workflow and pays for itself fast. Below we compare features, pricing, and real output so you can choose with confidence.
Read more →
ImageNet

The ImageNet project is a large visual database designed for use in visual object recognition software research. More than 14 million images have been hand-annotated by the project to indicate what objects are pictured and in at least one million of the images, bounding boxes are also provided. ImageNet contains more than 20,000 categories, with a typical category, such as "balloon" or "strawberry", consisting of several hundred images. The database of annotations of third-party image URLs is freely available directly from ImageNet, though the actual images are not owned by ImageNet. Since 2010, the ImageNet project runs an annual software contest, the ImageNet Large Scale Visual Recognition Challenge (ILSVRC), where software programs compete to correctly classify and detect objects and scenes. The challenge uses a "trimmed" list of one thousand non-overlapping classes. == History == AI researcher Fei-Fei Li began working on the idea for ImageNet in 2006. At a time when most AI research focused on models and algorithms, Li wanted to expand and improve the data available to train AI algorithms. In 2007, Li met with Princeton professor Christiane Fellbaum, one of the creators of WordNet, to discuss the project. As a result of this meeting, Li went on to build ImageNet starting from the roughly 22,000 nouns of WordNet and using many of its features. She was also inspired by a 1987 estimate that the average person recognizes roughly 30,000 different kinds of objects. As an assistant professor at Princeton, Li assembled a team of researchers to work on the ImageNet project. They used Amazon Mechanical Turk to help with the classification of images. Labeling started in July 2008 and ended in April 2010. It took 49K workers from 167 countries filtering and labeling over 160M candidate images. They had enough budget to have each of the 14 million images labelled three times. The original plan called for 10,000 images per category, for 40,000 categories at 400 million images, each verified 3 times. They found that humans can classify at most 2 images/sec. At this rate, it was estimated to take 19 human-years of labor (without rest). They presented their database for the first time as a poster at the 2009 Conference on Computer Vision and Pattern Recognition (CVPR) in Florida, titled "ImageNet: A Preview of a Large-scale Hierarchical Dataset". The poster was reused at Vision Sciences Society 2009. In 2009, Alex Berg suggested adding object localization as a task. Li approached PASCAL Visual Object Classes contest in 2009 for a collaboration. It resulted in the subsequent ImageNet Large Scale Visual Recognition Challenge starting in 2010, which has 1000 classes and object localization, as compared to PASCAL VOC which had just 20 classes and 19,737 images (in 2010). === Significance for deep learning === On 30 September 2012, a convolutional neural network (CNN) called AlexNet achieved a top-5 error of 15.3% in the ImageNet 2012 Challenge, more than 10.8 percentage points lower than that of the runner-up. Using convolutional neural networks was feasible due to the use of graphics processing units (GPUs) during training, an essential ingredient of the deep learning revolution. According to The Economist, "Suddenly people started to pay attention, not just within the AI community but across the technology industry as a whole." In 2015, AlexNet was outperformed by Microsoft's very deep CNN with over 100 layers, which won the ImageNet 2015 contest, having 3.57% error on the test set. Andrej Karpathy estimated in 2014 that with concentrated effort, he could reach 5.1% error rate, and ~10 people from his lab reached ~12-13% with less effort. It was estimated that with maximal effort, a human could reach 2.4%. == Dataset == ImageNet crowdsources its annotation process. Image-level annotations indicate the presence or absence of an object class in an image, such as "there are tigers in this image" or "there are no tigers in this image". Object-level annotations provide a bounding box around the (visible part of the) indicated object. ImageNet uses a variant of the broad WordNet schema to categorize objects, augmented with 120 categories of dog breeds to showcase fine-grained classification. In 2012, ImageNet was the world's largest academic user of Mechanical Turk. The average worker identified 50 images per minute. The original plan of the full ImageNet would have roughly 50M clean, diverse and full resolution images spread over approximately 50K synsets. This was not achieved. The summary statistics given on April 30, 2010: Total number of non-empty synsets: 21841 Total number of images: 14,197,122 Number of images with bounding box annotations: 1,034,908 Number of synsets with SIFT features: 1000 Number of images with SIFT features: 1.2 million === Categories === The categories of ImageNet were filtered from the WordNet concepts. Each concept, since it can contain multiple synonyms (for example, "kitty" and "young cat"), so each concept is called a "synonym set" or "synset". There were more than 100,000 synsets in WordNet 3.0, majority of them are nouns (80,000+). The ImageNet dataset filtered these to 21,841 synsets that are countable nouns that can be visually illustrated. Each synset in WordNet 3.0 has a "WordNet ID" (wnid), which is a concatenation of part of speech and an "offset" (a unique identifying number). Every wnid starts with "n" because ImageNet only includes nouns. For example, the wnid of synset "dog, domestic dog, Canis familiaris" is "n02084071". The categories in ImageNet fall into 9 levels, from level 1 (such as "mammal") to level 9 (such as "German shepherd"). === Image format === The images were scraped from online image search (Google, Picsearch, MSN, Yahoo, Flickr, etc) using synonyms in multiple languages. For example: German shepherd, German police dog, German shepherd dog, Alsatian, ovejero alemán, pastore tedesco, 德国牧羊犬. ImageNet consists of images in RGB format with varying resolutions. For example, in ImageNet 2012, "fish" category, the resolution ranges from 4288 x 2848 to 75 x 56. In machine learning, these are typically preprocessed into a standard constant resolution, and whitened, before further processing by neural networks. For example, in PyTorch, ImageNet images are by default normalized by dividing the pixel values so that they fall between 0 and 1, then subtracting by [0.485, 0.456, 0.406], then dividing by [0.229, 0.224, 0.225]. These are the mean and standard deviations for ImageNet, so this whitens the input data. === Labels and annotations === Each image is labelled with exactly one wnid. Dense SIFT features (raw SIFT descriptors, quantized codewords, and coordinates of each descriptor/codeword) for ImageNet-1K were available for download, designed for bag of visual words. The bounding boxes of objects were available for about 3000 popular synsets with on average 150 images in each synset. Furthermore, some images have attributes. They released 25 attributes for ~400 popular synsets: Color: black, blue, brown, gray, green, orange, pink, red, violet, white, yellow Pattern: spotted, striped Shape: long, round, rectangular, square Texture: furry, smooth, rough, shiny, metallic, vegetation, wooden, wet === ImageNet-21K === The full original dataset is referred to as ImageNet-21K. ImageNet-21k contains 14,197,122 images divided into 21,841 classes. Some papers round this up and name it ImageNet-22k. The full ImageNet-21k was released in Fall of 2011, as fall11_whole.tar. There is no official train-validation-test split for ImageNet-21k. Some classes contain only 1-10 samples, while others contain thousands. === ImageNet-1K === There are various subsets of the ImageNet dataset used in various context, sometimes referred to as "versions". One of the most highly used subsets of ImageNet is the "ImageNet Large Scale Visual Recognition Challenge (ILSVRC) 2012–2017 image classification and localization dataset". This is also referred to in the research literature as ImageNet-1K or ILSVRC2017, reflecting the original ILSVRC challenge that involved 1,000 classes. ImageNet-1K contains 1,281,167 training images, 50,000 validation images and 100,000 test images. Each category in ImageNet-1K is a leaf category, meaning that there are no child nodes below it, unlike ImageNet-21K. For example, in ImageNet-21K, there are some images categorized as simply "mammal", whereas in ImageNet-1K, there are only images categorized as things like "German shepherd", since there are no child-words below "German shepherd". === Later developments === In the WordNet they built ImageNet on, there were 2832 synsets in the "person" subtree. During 2018--2020 period, they removed the download of the ImageNet-21k as they went through extensive filtering in these person synsets. Out of these 2832 synsets, 1593 were deemed "potentially offensive". Out of the remaining 1239, 1081 were deemed not really "visual". The result was that only 158 syn
Read more →
Katz's back-off model

Katz back-off is a generative n-gram language model that estimates the conditional probability of a word given its history in the n-gram. It accomplishes this estimation by backing off through progressively shorter history models under certain conditions. By doing so, the model with the most reliable information about a given history is used to provide the better results. The model was introduced in 1987 by Slava M. Katz. Prior to that, n-gram language models were constructed by training individual models for different n-gram orders using maximum likelihood estimation and then interpolating them together. == Method == The equation for Katz's back-off model is: P b o ( w i ∣ w i − n + 1 ⋯ w i − 1 ) = { d w i − n + 1 ⋯ w i C ( w i − n + 1 ⋯ w i − 1 w i ) C ( w i − n + 1 ⋯ w i − 1 ) if C ( w i − n + 1 ⋯ w i ) > k α w i − n + 1 ⋯ w i − 1 P b o ( w i ∣ w i − n + 2 ⋯ w i − 1 ) otherwise {\displaystyle {\begin{aligned}&P_{bo}(w_{i}\mid w_{i-n+1}\cdots w_{i-1})\\[4pt]={}&{\begin{cases}d_{w_{i-n+1}\cdots w_{i}}{\dfrac {C(w_{i-n+1}\cdots w_{i-1}w_{i})}{C(w_{i-n+1}\cdots w_{i-1})}}&{\text{if }}C(w_{i-n+1}\cdots w_{i})>k\\[10pt]\alpha _{w_{i-n+1}\cdots w_{i-1}}P_{bo}(w_{i}\mid w_{i-n+2}\cdots w_{i-1})&{\text{otherwise}}\end{cases}}\end{aligned}}} where C(x) = number of times x appears in training wi = ith word in the given context Essentially, this means that if the n-gram has been seen more than k times in training, the conditional probability of a word given its history is proportional to the maximum likelihood estimate of that n-gram. Otherwise, the conditional probability is equal to the back-off conditional probability of the (n − 1)-gram. The more difficult part is determining the values for k, d and α. k {\displaystyle k} is the least important of the parameters. It is usually chosen to be 0. However, empirical testing may find better values for k. d {\displaystyle d} is typically the amount of discounting found by Good–Turing estimation. In other words, if Good–Turing estimates C {\displaystyle C} as C ∗ {\displaystyle C^{}} , then d = C ∗ C {\displaystyle d={\frac {C^{}}{C}}} To compute α {\displaystyle \alpha } , it is useful to first define a quantity β, which is the left-over probability mass for the (n − 1)-gram: β w i − n + 1 ⋯ w i − 1 = 1 − ∑ { w i : C ( w i − n + 1 ⋯ w i ) > k } d w i − n + 1 ⋯ w i C ( w i − n + 1 ⋯ w i − 1 w i ) C ( w i − n + 1 ⋯ w i − 1 ) {\displaystyle \beta _{w_{i-n+1}\cdots w_{i-1}}=1-\sum _{\{w_{i}:C(w_{i-n+1}\cdots w_{i})>k\}}d_{w_{i-n+1}\cdots w_{i}}{\frac {C(w_{i-n+1}\cdots w_{i-1}w_{i})}{C(w_{i-n+1}\cdots w_{i-1})}}} Then the back-off weight, α, is computed as follows: α w i − n + 1 ⋯ w i − 1 = β w i − n + 1 ⋯ w i − 1 ∑ { w i : C ( w i − n + 1 ⋯ w i ) ≤ k } P b o ( w i ∣ w i − n + 2 ⋯ w i − 1 ) {\displaystyle \alpha _{w_{i-n+1}\cdots w_{i-1}}={\frac {\beta _{w_{i-n+1}\cdots w_{i-1}}}{\sum _{\{w_{i}:C(w_{i-n+1}\cdots w_{i})\leq k\}}P_{bo}(w_{i}\mid w_{i-n+2}\cdots w_{i-1})}}} The above formula only applies if there is data for the "(n − 1)-gram". If not, the algorithm skips n-1 entirely and uses the Katz estimate for n-2. (and so on until an n-gram with data is found) == Discussion == This model generally works well in practice, but fails in some circumstances. For example, suppose that the bigram "a b" and the unigram "c" are very common, but the trigram "a b c" is never seen. Since "a b" and "c" are very common, it may be significant (that is, not due to chance) that "a b c" is never seen. Perhaps it's not allowed by the rules of the grammar. Instead of assigning a more appropriate value of 0, the method will back off to the bigram and estimate P(c | b), which may be too high.
Read more →
How to Choose an AI Content Generator

Curious about the best AI content generator? An AI content generator is software that uses machine learning to help you get more done — it combines speed, accuracy, and an interface that just works. Hands-on testing shows real-world results vary, so a short free trial is the smartest way to decide. Whether you are a beginner or a pro, the right AI content generator slots into your workflow and pays for itself fast. Read on for hands-on impressions, pricing tiers, and the standout features that matter.
Read more →
Jiaya Jia

Jiaya Jia (Chinese: 贾佳亚) is a Chair Professor of the Department of Computer Science and Engineering at The Hong Kong University of Science and Technology (HKUST). He is an IEEE Fellow, the associate editor-in-chief of one of IEEE’s flagship and premier journals- Transactions on Pattern Analysis and Machine Intelligence (TPAMI), as well as on the editorial board of International Journal of Computer Vision (IJCV). == Early life and education == Jiaya Jia joined CUHK in 2004 as an assistant professor, and was promoted to full professor in 2015. He obtained his PhD degree in computer science jointly from Hong Kong University of Science and Technology and Microsoft Research in 2004. From March 2003 to August 2004, he was a visiting scholar at Microsoft. He conducted collaborative research at Adobe Research in 2007. == Career == Jiaya Jia is a distinguished scientist in the fields of computer vision and artificial intelligence. His research team at HKUST, DV Lab, is one of the largest vision AI research teams in the world and has been making significant contribution to advanced development of computer vision algorithms and technologies with focuses on image/video understanding, detection and segmentation, multi-modal AI, computational imaging, practical optimization, and advanced learning for visual content since 2000. Jiaya Jia has published 200+ top papers and was cited 80,000+ times on Google Scholar with H-Index 110+. 40+ PhDs and fellows from this group are now active in academia and industry, and have become prominent AI tech leaders as professors, directors in major research labs, and founders of several successful startups. Jiaya Jia assumes the position of associate editor-in-chief of IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI) since 2021. He is also on the editorial board of International Journal of Computer Vision (IJCV). Jiaya Jia has served as the area chair of ICCV, CVPR, AAAI, ECCV, and several other premium international AI conferences for years. He was on program committees of major conferences in graphics and computational imaging, including ICCP, SIGGRAPH, and SIGGRAPH Asia. == Research == The research areas of Jiaya Jia are computer vision, large X models, and deep learning. Jiaya Jia has made outstanding contributions to computer vision technology, algorithms and engineering, and is among the world's leading experts in the field. His research partners include numerous renowned multinational technology companies, such as Microsoft, Qualcomm, Adobe, Intel, NVIDIA, Amazon, and Lenovo. Jia has cultivated a number of outstanding talents with Master's and PhDs who continue to engage in scientific research and development in computer vision. Many technologies in image analysis and processing developed by Jiaya Jia are still leading in the field worldwide. Wherein, his achievements in image deblurring, filtering, image sparse processing, multi-band image signal fusion and enhancement, large range motion estimation, texture and structure-based layering, etc. have been published in the industry's most influential conferences and publications, and implemented in the real-world applications. These achievements have demonstrated outstanding performance in established systems, and most of which are open source so as to enable wider applications across industries such as aviation, medical imaging, safety management, robotic design, meteorological analysis and many more. == Selected publications == In his over 20 years of research experience, Jiaya Jia has published 200+ top papers that have been cited more than 80,000 times. According to HKUST Website in August 2024, Jiaya Jia has accumulatively published over 200 scientific papers in books, journals and conferences, such as IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), International Journal of Computer Vision (IJCV) "Computer Vision and Pattern Recognition (CVPR)", and "International Conference on Computer Vision (ICCV)". Representative papers include: Jiaya Jia: Mathematical Models and Practical Solvers for Uniform Motion Deblurring (in Motion Deblurring: Algorithms and Systems), Cambridge University Press, ISBN 9781107044364, 2014; Jiaya Jia: “Matte Extraction” Book: Computer Vision - A Reference Guide, Springer, ISBN 9780387307718 Editor-in-chief: Ikeuchi, Katsushi; Jiaya Jia, Chi-Keung Tang：Image Stitching Using Structure Deformation，IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), Vol. 30, No. 4, 2008; Jiaya Jia, Jian Sun, Chi-Keung Tang, Heung-Yeung Shum：Drag-and-Drop Pasting，ACM Transactions on Graphics (also in SIGGRAPH 2006), Vol. 25, No. 3, 2006. Xiaojuan Qi, Zheng zhe Liu, Renjie Liao, Philip HS Torr, Raquel Urtasun, Jiaya Jia：GeoNet++: Iterative Geometric Neural Network with Edge-Aware Refinement for Joint Depth and Surface Normal Estimation，IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI). Accepted. == Selected honors and awards == ACM Fellow. 1st Place of WAD Drivable Area Segmentation Challenge 2018; 1st Place of LSUN'17 Instance and Semantic Segmentation Challenges; 1st Place of COCO Instance Segmentation Challenge 2017; 2nd Place in COCO Detection Challenge 2017; 1st Place of ImageNet Scene Parsing Challenge 2016 with the paper PSPNet presented in CVPR 2017.
Read more →
Kernel density estimation

In statistics, kernel density estimation (KDE) is the application of kernel smoothing for probability density estimation, i.e., a non-parametric method to estimate the probability density function of a random variable based on kernels as weights. KDE answers a fundamental data smoothing problem where inferences about the population are made based on a finite data sample. In some fields such as signal processing and econometrics it is also termed the Parzen–Rosenblatt window method, after Emanuel Parzen and Murray Rosenblatt, who are usually credited with independently creating it in its current form. One of the famous applications of kernel density estimation is in estimating the class-conditional marginal densities of data when using a naive Bayes classifier, which can improve its prediction accuracy. == Definition == Let x = ( x 1 , x 2 , x 3 , . . . ) {\displaystyle \mathbf {x} =\left(x_{1},x_{2},x_{3},...\right)} be independent and identically distributed samples drawn from some univariate distribution with an unknown density f at any given point x. We are interested in estimating the shape of this function f. Its kernel density estimator is f ^ h ( x ) = 1 n ∑ i = 1 n K h ( x − x i ) = 1 n h ∑ i = 1 n K ( x − x i h ) , {\displaystyle {\hat {f}}_{h}(x)={\frac {1}{n}}\sum _{i=1}^{n}K_{h}(x-x_{i})={\frac {1}{nh}}\sum _{i=1}^{n}K{\left({\frac {x-x_{i}}{h}}\right)},} where K is the kernel — a non-negative function — and h > 0 is a smoothing parameter called the bandwidth or simply width. A kernel with subscript h is called the scaled kernel and defined as Kh(x) = ⁠1/h⁠ K(⁠x/h⁠). Intuitively one wants to choose h as small as the data will allow; however, there is always a trade-off between the bias of the estimator and its variance. The choice of bandwidth is discussed in more detail below. A range of kernel functions are commonly used: uniform, triangular, biweight, triweight, Epanechnikov (parabolic), normal, and others. The Epanechnikov kernel is optimal in a mean square error sense, though the loss of efficiency is small for the kernels listed previously. Due to its convenient mathematical properties, the normal kernel is often used, which means K(x) = ϕ(x), where ϕ is the standard normal density function. The kernel density estimator then becomes f ^ h ( x ) = 1 n ∑ i = 1 n 1 h 2 π exp ⁡ ( − ( x − x i ) 2 2 h 2 ) , {\displaystyle {\hat {f}}_{h}(x)={\frac {1}{n}}\sum _{i=1}^{n}{\frac {1}{h{\sqrt {2\pi }}}}\exp \left({\frac {-(x-x_{i})^{2}}{2h^{2}}}\right),} where h {\displaystyle h} is the standard deviation of the sample x {\displaystyle \mathbf {x} } . The construction of a kernel density estimate finds interpretations in fields outside of density estimation. For example, in thermodynamics, this is equivalent to the amount of heat generated when heat kernels (the fundamental solution to the heat equation) are placed at each data point locations xi. Similar methods are used to construct discrete Laplace operators on point clouds for manifold learning (e.g. diffusion map). == Example == Kernel density estimates are closely related to histograms, but can be endowed with properties such as smoothness or continuity by using a suitable kernel. The diagram below based on these 6 data points illustrates this relationship: For the histogram, first, the horizontal axis is divided into sub-intervals or bins which cover the range of the data: In this case, six bins each of width 2. Whenever a data point falls inside this interval, a box of height 1/12 is placed there. If more than one data point falls inside the same bin, the boxes are stacked on top of each other. For the kernel density estimate, normal kernels with a standard deviation of 1.5 (indicated by the red dashed lines) are placed on each of the data points xi. The kernels are summed to make the kernel density estimate (solid blue curve). The smoothness of the kernel density estimate (compared to the discreteness of the histogram) illustrates how kernel density estimates converge faster to the true underlying density for continuous random variables. == Bandwidth selection == The bandwidth of the kernel is a free parameter which exhibits a strong influence on the resulting estimate. To illustrate its effect, we take a simulated random sample from the standard normal distribution (plotted at the blue spikes in the rug plot on the horizontal axis). The grey curve is the true density (a normal density with mean 0 and variance 1). In comparison, the red curve is undersmoothed since it contains too many spurious data artifacts arising from using a bandwidth h = 0.05, which is too small. The green curve is oversmoothed since using the bandwidth h = 2 obscures much of the underlying structure. The black curve with a bandwidth of h = 0.337 is considered to be optimally smoothed since its density estimate is close to the true density. An extreme situation is encountered in the limit h → 0 {\displaystyle h\to 0} (no smoothing), where the estimate is a sum of n delta functions centered at the coordinates of analyzed samples. In the other extreme limit h → ∞ {\displaystyle h\to \infty } the estimate retains the shape of the used kernel, centered on the mean of the samples (completely smooth). The most common optimality criterion used to select this parameter is the expected L2 risk function, also termed the mean integrated squared error: MISE ⁡ ( h ) = E [ ∫ ( f ^ h ( x ) − f ( x ) ) 2 d x ] {\displaystyle \operatorname {MISE} (h)=\operatorname {E} \!\left[\int \!{\left({\hat {f}}\!_{h}(x)-f(x)\right)}^{2}dx\right]} Under weak assumptions on f and K, (f is the, generally unknown, real density function), MISE ⁡ ( h ) = AMISE ⁡ ( h ) + o ( ( n h ) − 1 + h 4 ) {\displaystyle \operatorname {MISE} (h)=\operatorname {AMISE} (h)+{\mathcal {o}}{\left((nh)^{-1}+h^{4}\right)}} where o is the little o notation, and n the sample size (as above). The AMISE is the asymptotic MISE, i. e. the two leading terms, AMISE ⁡ ( h ) = R ( K ) n h + 1 4 m 2 ( K ) 2 h 4 R ( f ″ ) {\displaystyle \operatorname {AMISE} (h)={\frac {R(K)}{nh}}+{\frac {1}{4}}m_{2}(K)^{2}h^{4}R(f'')} where R ( g ) = ∫ g ( x ) 2 d x {\textstyle R(g)=\int g(x)^{2}\,dx} for a function g, m 2 ( K ) = ∫ x 2 K ( x ) d x {\textstyle m_{2}(K)=\int x^{2}K(x)\,dx} and f ″ {\displaystyle f''} is the second derivative of f {\displaystyle f} and K {\displaystyle K} is the kernel. The minimum of this AMISE is the solution to this differential equation ∂ ∂ h AMISE ⁡ ( h ) = − R ( K ) n h 2 + m 2 ( K ) 2 h 3 R ( f ″ ) = 0 {\displaystyle {\frac {\partial }{\partial h}}\operatorname {AMISE} (h)=-{\frac {R(K)}{nh^{2}}}+m_{2}(K)^{2}h^{3}R(f'')=0} or h AMISE = R ( K ) 1 / 5 m 2 ( K ) 2 / 5 R ( f ″ ) 1 / 5 n − 1 / 5 = C n − 1 / 5 {\displaystyle h_{\operatorname {AMISE} }={\frac {R(K)^{1/5}}{m_{2}(K)^{2/5}R(f'')^{1/5}}}n^{-1/5}=Cn^{-1/5}} Neither the AMISE nor the hAMISE formulas can be used directly since they involve the unknown density function f {\displaystyle f} or its second derivative f ″ {\displaystyle f''} . To overcome that difficulty, a variety of automatic, data-based methods have been developed to select the bandwidth. Several review studies have been undertaken to compare their efficacies, with the general consensus that the plug-in selectors and cross validation selectors are the most useful over a wide range of data sets. Substituting any bandwidth h which has the same asymptotic order n−1/5 as hAMISE into the AMISE gives that AMISE(h) = O(n−4/5), where O is the big O notation. It can be shown that, under weak assumptions, there cannot exist a non-parametric estimator that converges at a faster rate than the kernel estimator. Note that the n−4/5 rate is slower than the typical n−1 convergence rate of parametric methods. If the bandwidth is not held fixed, but is varied depending upon the location of either the estimate (balloon estimator) or the samples (pointwise estimator), this produces a particularly powerful method termed adaptive or variable bandwidth kernel density estimation. Bandwidth selection for kernel density estimation of heavy-tailed distributions is relatively difficult. === A rule-of-thumb bandwidth estimator === If Gaussian basis functions are used to approximate univariate data, and the underlying density being estimated is Gaussian, the optimal choice for h (that is, the bandwidth that minimises the mean integrated squared error) is: h = ( 4 σ ^ 5 3 n ) 1 / 5 ≈ 1.06 σ ^ n − 1 / 5 , {\displaystyle h={\left({\frac {4{\hat {\sigma }}^{5}}{3n}}\right)}^{1/5}\approx 1.06\,{\hat {\sigma }}\,n^{-1/5},} An h {\displaystyle h} value is considered more robust when it improves the fit for long-tailed and skewed distributions or for bimodal mixture distributions. This is often done empirically by replacing the standard deviation σ ^ {\displaystyle {\hat {\sigma }}} by the parameter A {\displaystyle A} below: A = min ( σ ^ , I Q R 1.34 ) {\displaystyle A=\min \left({\hat {\sigma }},{\frac {\mathrm {IQR} }{1.34}}\right)} where IQR is the
Read more →
Geoffrey J. Gordon

Geoffrey J. Gordon is a professor at the Machine Learning Department at Carnegie Mellon University in Pittsburgh and director of research at the Microsoft Montréal lab. He is known for his research in statistical relational learning (a subdiscipline of artificial intelligence and machine learning) and on anytime dynamic variants of the A search algorithm. His research interests include multi-agent planning, reinforcement learning, decision-theoretic planning, statistical models of difficult data (e.g. maps, video, text), computational learning theory, and game theory. Gordon received a B.A. in computer science from Cornell University in 1991, and a PhD at Carnegie Mellon in 1999.
Read more →
Glushkov's construction algorithm

In computer science theory – particularly formal language theory – Glushkov's construction algorithm, invented by Victor Mikhailovich Glushkov, transforms a given regular expression into an equivalent nondeterministic finite automaton (NFA). Thus, it forms a bridge between regular expressions and nondeterministic finite automata: two abstract representations of the same class of formal languages. A regular expression may be used to conveniently describe an advanced search pattern in a "find and replace"–like operation of a text processing utility. Glushkov's algorithm can be used to transform it into an NFA, which furthermore is small by nature, as the number of its states equals the number of symbols of the regular expression, plus one. Subsequently, the NFA can be made deterministic by the powerset construction and then be minimized to get an optimal automaton corresponding to the given regular expression. The latter format is best suited for execution on a computer. From another, more theoretical point of view, Glushkov's algorithm is a part of the proof that NFA and regular expressions both accept exactly the same languages; that is, the regular languages. The converse of Glushkov's algorithm is Kleene's algorithm, which transforms a finite automaton into a regular expression. The automaton obtained by Glushkov's construction is the same as the one obtained by Thompson's construction algorithm, once its ε-transitions are removed. Glushkov's construction algorithm is also called The algorithm of Berry-Sethi, named after Gérard Berry and Ravi Sethi who worked on this construction. == Construction == Given a regular expression e, the Glushkov Construction Algorithm creates a non-deterministic automaton that accepts the language L ( e ) {\displaystyle L(e)} accepted by e. The construction uses four steps: === Step 1 === Linearisation of the expression. Each letter of the alphabet appearing in the expression e is renamed, so that each letter occurs at most once in the new expression e ′ {\displaystyle e'} . Glushkov's construction essentially relies on the fact that e ′ {\displaystyle e'} represents a local language L ( e ′ ) {\displaystyle L(e')} . Let A be the old alphabet and let B be the new one. === Step 2a === Computation of the sets P ( e ′ ) {\displaystyle P(e')} , D ( e ′ ) {\displaystyle D(e')} , and F ( e ′ ) {\displaystyle F(e')} . The first, P ( e ′ ) {\displaystyle P(e')} , is the set of letters which occurs as first letter of a word of L ( e ′ ) {\displaystyle L(e')} . The second, D ( e ′ ) {\displaystyle D(e')} , is the set of letters that can end a word of L ( e ′ ) {\displaystyle L(e')} . The last one, F ( e ′ ) {\displaystyle F(e')} , is the set of letter pairs that can occur in words of L ( e ′ ) {\displaystyle L(e')} , i.e. it is the set of factors of length two of the words of L ( e ′ ) {\displaystyle L(e')} . Those sets are mathematically defined by P ( e ′ ) = { x ∈ B ∣ x B ∗ ∩ L ( e ′ ) ≠ ∅ } {\displaystyle P(e')=\{x\in B\mid xB^{}\cap L(e')\neq \emptyset \}} , D ( e ′ ) = { y ∈ B ∣ B ∗ y ∩ L ( e ′ ) ≠ ∅ } {\displaystyle D(e')=\{y\in B\mid B^{}y\cap L(e')\neq \emptyset \}} , F ( e ′ ) = { u ∈ B 2 ∣ B ∗ u B ∗ ∩ L ( e ′ ) ≠ ∅ } {\displaystyle F(e')=\{u\in B^{2}\mid B^{}uB^{}\cap L(e')\neq \emptyset \}} . They are computed by induction over the structure of the expression, as explained below. === Step 2b === Computation of the set Λ ( e ′ ) {\displaystyle \Lambda (e')} which contains the empty word ε {\displaystyle \varepsilon } if this word belongs to L ( e ′ ) {\displaystyle L(e')} , and is the empty set otherwise. Formally, this is Λ ( e ′ ) = { ε } ∩ L ( e ′ ) {\displaystyle \Lambda (e')=\{\varepsilon \}\cap L(e')} . === Step 3 === Computation of automaton recognizing the local language, as defined by P ( e ′ ) {\displaystyle P(e')} , D ( e ′ ) {\displaystyle D(e')} , F ( e ′ ) {\displaystyle F(e')} , and Λ ( e ′ ) {\displaystyle \Lambda (e')} . By definition, the local language defined by the sets P, D, and F is the set of words which begin with a letter of P, end by a letter of D, and whose factors of length 2 belong to F, optionally also including the empty word; that is, it is the language: L ′ = ( P B ∗ ∩ B ∗ D ) ∖ B ∗ ( B 2 ∖ F ) B ∗ ∪ Λ ( e ′ ) {\displaystyle L'=(PB^{}\cap B^{}D)\setminus B^{}(B^{2}\setminus F)B^{}\cup \Lambda (e')} . Strictly speaking, it is the computation of the automaton for the local language denoted by this linearised expression that is Glushkov's construction. === Step 4 === Remove the linearisation, replacing each indexed letter B by the original letter of A. == Example == Consider the regular expression e = ( a ( a b ) ∗ ) ∗ + ( b a ) ∗ {\displaystyle e=(a(ab)^{})^{}+(ba)^{}} . == Computation of the set of letters == The computation of the sets P, D, F, and Λ is done inductively over the regular expression e ′ {\displaystyle e'} . One must give the values for ∅, ε (the symbols for the empty language and the singleton language containing the empty word), the letters, and the results of the operations + , ⋅ , ∗ {\displaystyle +,\cdot ,^{}} . The most costly operations are the cartesian products of sets for the computation of F. == Properties == The obtained automaton is non-deterministic, and it has as many states as the number of letters of the regular expression, plus one. It has been proven that every Thompson's automaton can be transformed into Glushkov's automaton via a ε-transitions elimination method. == Applications and deterministic expressions == The computation of the automaton by the expression occurs often; it has been systematically used in search functions, in particular by the Unix grep command. Similarly, XML's specification also uses such constructions; for more efficiency, regular expressions of a certain kind, called deterministic expressions, have been studied.
Read more →
James F. Allen (computer scientist)

James Frederick Allen (born 1950) is an American computational linguist recognized for his contributions to temporal logic, in particular Allen's interval algebra. He is interested in knowledge representation, commonsense reasoning, and natural language understanding, believing that "deep language understanding can only currently be achieved by significant hand-engineering of semantically-rich formalisms coupled with statistical preferences". He is the John H. Dessaurer Professor of Computer Science at the University of Rochester. == Biography == Allen received his Ph.D. from the University of Toronto in 1979, under the supervision of C. Raymond Perrault, after which he joined the faculty at Rochester. At Rochester, he was department chair from 1987 to 1990, directed the Cognitive Science Program from 1992 to 1996, and co-directed the Center for the Sciences of Language from 1996 to 1998. He served as the Editor-in-Chief of Computational Linguistics from 1983–1993. Since 2006 he has also been associate director of the Florida Institute for Human and Machine Cognition. == Academic life == === TRIPS project === The TRIPS project is a long-term research to build generic technology for dialogue (both spoken and 'chat') systems, which includes natural language processing, collaborative problem solving, and dynamic context-sensitive language modeling. This is contrast with the data driven approaches by machine learning, which requires to collect and annotate corpora, i.e. training data, firstly. === PLOW agent === PLOW agent is a system that learns executable task models from a single collaborative learning session, which integrates wide AI technologies including deep natural language understanding, knowledge representation and reasoning, dialogue systems, planning/agent-based systems, and machine learning. This paper won the outstanding paper award at AAAI in 2007. == Selected works == === Books === Allen is the author of the textbook Natural Language Understanding (Benjamin-Cummings, 1987; 2nd ed., 1995). He is also the co-author with Henry Kautz, Richard Pelavin, and Josh Tenenberg of Reasoning About Plans (Morgan Kaufmann, 1991). === Articles === 2007. PLOW: A Collaborative Task Learning Agent. (with Nathanael Chambers et al) AAAI'07 won the outstanding paper award at AAAI in 2007. 2006. Chester: Towards a Personal Medication Advisor. (with N. Blaylock, et al) Biomedical informatics 39(5) 1998. TRIPS: An Integrated Intelligent Problem-Solving Assistant. (with George Ferguson) AAAI'98 1983. Maintaining Knowledge about Temporal Intervals. CACM 26, 11, 832-843 == Awards and honors == In 1991 he was elected as a fellow of the Association for the Advancement of Artificial Intelligence (1990, founding fellow). In 1992 he became the Dessaurer Professor at Rochester.
Read more →
Google Vids

Google Vids (not to be confused with Google Video) is an online timeline-based video editing application included as part of the Google Workspace suite. It is designed to help users create informational videos for work-related purposes. The app uses Google's Gemini technology to enable users to create video storyboards manually or with AI assistance using simple prompts. Features include uploading media, choosing stock videos, images, background music, and a voiceover feature with script generation using AI. The app is currently in testing with select Google Workspace Labs users. Like Kapwing and Capcut, Google Vids is primarily for creating work-related content like sales training, onboarding videos, vendor outreach, and project updates. It offers various styles and templates, collaborative features, and is not limited to videos without the short integration at the moment. Google Vids was announced on April 9, 2024. In September 2025, Google began to roll out a basic version of the application to Google Workspace users.
Read more →
Michael Kearns (computer scientist)

Michael Justin Kearns is an American computer scientist, professor and National Center Chair at the University of Pennsylvania, the founding director of Penn's Singh Program in Networked & Social Systems Engineering (NETS), the founding director of Warren Center for Network and Data Sciences, and also holds secondary appointments in Penn's Wharton School and department of Economics. He is a leading researcher in computational learning theory and algorithmic game theory, and interested in machine learning, artificial intelligence, computational finance, algorithmic trading, computational social science and social networks. He previously led the Advisory and Research function in Morgan Stanley's Artificial Intelligence Center of Excellence team, and is currently an Amazon Scholar within Amazon Web Services. == Biography == Kearns was born into an academic family, where his father David R Kearns is Professor Emeritus at University of California, San Diego in chemistry, who won Guggenheim Fellowship in 1969, and his uncle Thomas R. Kearns is Professor Emeritus at Amherst College in Philosophy and Law, Jurisprudence, and Social Thought. His paternal grandfather Clyde W. Kearns was a pioneer in insecticide toxicology and was a professor at University of Illinois at Urbana–Champaign in Entomology, and his maternal grandfather Chen Shou-Yi (1899–1978) was a professor at Pomona College in history and literature, who was born in Canton (Guangzhou, China) into a family noted for their scholarship and educational leadership. Kearns received his B.S. degree at the University of California at Berkeley in math and computer science in 1985, and Ph.D. in computer science from Harvard University in 1989, under the supervision of Turing Award winner Leslie Valiant. His doctoral dissertation was The Computational Complexity of Machine Learning, later published by MIT press as part of the ACM Doctoral Dissertation Award Series in 1990. Before joining AT&T Bell Labs in 1991, he continued with postdoctoral positions at the Laboratory for Computer Science at MIT hosted by Ronald Rivest, and at the International Computer Science Institute (ICSI) in UC Berkeley hosted by Richard M. Karp, both of whom are Turing Award winners. Kearns is currently a full professor and National Center Chair at the University of Pennsylvania, where his appointment is split across the Department of Computer and Information Science, and Statistics and Operations and Information Management in the Wharton School. Prior to joining the Penn faculty in 2002, he spent a decade (1991–2001) in AT&T Labs and Bell Labs, including as head of the AI department with colleagues including Michael L. Littman, David A. McAllester, and Richard S. Sutton; Secure Systems Research department; and Machine Learning department with members such as Michael Collins and the leader Fernando Pereira. Other AT&T Labs colleagues in Algorithms and Theoretical Computer Science included Yoav Freund, Ronald Graham, Mehryar Mohri, Robert Schapire, and Peter Shor, as well as Sebastian Seung, Yann LeCun, Corinna Cortes, and Vladimir Vapnik (the V in VC dimension). Kearns was named Fellow of the Association for Computing Machinery (2014) for contributions to machine learning, and a fellow of the American Academy of Arts and Sciences (2012). His former graduate students and postdoctoral visitors include Ryan W. Porter, John Langford, and Jennifer Wortman Vaughan. Kearns' work has been reported by media, such as MIT Technology Review (2014) Can a Website Help You Decide to Have a Kid?, Bloomberg News (2014) Schneiderman (and Einstein) Pressure High-Speed Trading and NPR audio (2012) Online Education Grows Up, And For Now, It's Free. == Academic life == === Computational learning theory === Kearns and Umesh Vazirani published An introduction to computational learning theory, which has been a standard text on computational learning theory since it was published in 1994. === Weak learnability and the origin of Boosting algorithms === The question "is weakly learnability equivalent to strong learnability?" posed by Kearns and Valiant (Unpublished manuscript 1988, ACM Symposium on Theory of Computing 1989) is the origin of boosting machine learning algorithms, which got a positive answer by Robert Schapire (1990, proof by construction, not practical) and Yoav Freund (1993, by voting, not practical) and then they developed the practical AdaBoost (European Conference on Computational Learning Theory 1995, Journal of Computer and System Sciences 1997), an adaptive boosting algorithm that won the prestigious Gödel Prize (2003). == Honors and awards == 2021. Member of the U. S. National Academy of Sciences. 2014. ACM Fellow. For contributions to machine learning, artificial intelligence, and algorithmic game theory and computational social science. 2012. American Academy of Arts and Sciences Fellow. == Selected works == 2019. The Ethical Algorithm: The Science of Socially Aware Algorithm Design. (with Aaron Roth). Oxford University Press. 1994. An introduction to computational learning theory. (with Umesh Vazirani). MIT press. Widely used as a text book in computational learning theory courses. 1990. The computational complexity of machine learning. MIT press. Based on his 1989 doctoral dissertation; ACM Doctoral Dissertation Award Series in 1990 Archived 2014-11-03 at the Wayback Machine 1989. Cryptographic limitations on learning Boolean formulae and finite automata. (with Leslie Valiant) Proceedings of the twenty-first annual ACM symposium on Theory of computing (STOC'89). The open question: is weakly learnability equivalent to strong learnability?; The origin of boosting algorithms; Important publication in machine learning.
Read more →
Amebis

Amebis from Kamnik is a company in Slovenia in the field of language technologies. The company has published several electronic dictionaries and encyclopedic dictionaries (e.g. ASP (32) dictionaries) and developed spell checkers, grammar checker Besana, hyphenators and lemmatizers for Slovene, Serbian and Albanian languages. The company maintains and edits the largest Slovenian dictionary portal Termania, which contains more than 135 dictionaries. The most used terminological dictionary on Termania is the Slovenian medical dictionary. In co-operation with company Alpineon and the Jožef Stefan Institute they have developed a speech synthesizer and screen reader Govorec (Speaker). They have also provided technical support for the largest text corpus of Slovene, called FidaPLUS, Fran and Franček. Amebis also developed the system of machine translation Amebis Presis, which incorporates the Slovenian language. On 11 October 2023 Amebis received award of the Father Stanislav Škrabec Foundation for special achievements in Slovene linguistics.
Read more →