AI Apple

AI Apple — independent reviews, comparisons, pricing and step-by-step guides on Aizhi.

Outline of machine learning

The following outline is provided as an overview of, and topical guide to, machine learning: Machine learning (ML) is a subfield of artificial intelligence within computer science that evolved from the study of pattern recognition and computational learning theory. In 1959, Arthur Samuel defined machine learning as a "field of study that gives computers the ability to learn without being explicitly programmed". ML involves the study and construction of algorithms that can learn from and make predictions on data. These algorithms operate by building a model from a training set of example observations to make data-driven predictions or decisions expressed as outputs, rather than following strictly static program instructions. == How can machine learning be categorized? == An academic discipline A branch of science An applied science A subfield of computer science A branch of artificial intelligence A subfield of soft computing Application of statistics === Paradigms of machine learning === Supervised learning, where the model is trained on labeled data Unsupervised learning, where the model tries to identify patterns in unlabeled data Reinforcement learning, where the model learns to make decisions by receiving rewards or penalties. == Applications of machine learning == Applications of machine learning Bioinformatics Biomedical informatics Computer vision Customer relationship management Data mining Earth sciences Email filtering Inverted pendulum (balance and equilibrium system) Natural language processing Named Entity Recognition Automatic summarization Automatic taxonomy construction Dialog system Grammar checker Language recognition Handwriting recognition Optical character recognition Speech recognition Text to Speech Synthesis Speech Emotion Recognition Machine translation Question answering Speech synthesis Text mining Term frequency–inverse document frequency Text simplification Pattern recognition Facial recognition system Handwriting recognition Image recognition Optical character recognition Speech recognition Recommendation system Collaborative filtering Content-based filtering Hybrid recommender systems Search engine Search engine optimization Social engineering == Machine learning hardware == Graphics processing unit Tensor processing unit Vision processing unit == Machine learning tools == Comparison of machine learning software Comparison of deep learning software === Machine learning frameworks === ==== Proprietary machine learning frameworks ==== Amazon Machine Learning Microsoft Azure Machine Learning Studio DistBelief (replaced by TensorFlow) ==== Open source machine learning frameworks ==== Apache Singa Apache MXNet Caffe PyTorch mlpack TensorFlow Torch CNTK Accord.Net Jax MLJ.jl – A machine learning framework for Julia === Machine learning libraries === Deeplearning4j Theano scikit-learn Keras === Machine learning algorithms === == Machine learning methods == === Instance-based algorithm === K-nearest neighbors algorithm (KNN) Learning vector quantization (LVQ) Self-organizing map (SOM) === Regression analysis === Logistic regression Ordinary least squares regression (OLSR) Linear regression Stepwise regression Multivariate adaptive regression splines (MARS) Regularization algorithm Ridge regression Least Absolute Shrinkage and Selection Operator (LASSO) Elastic net Least-angle regression (LARS) Classifiers Probabilistic classifier Naive Bayes classifier Binary classifier Linear classifier Hierarchical classifier === Dimensionality reduction === Dimensionality reduction Canonical correlation analysis (CCA) Factor analysis Feature extraction Feature selection Independent component analysis (ICA) Linear discriminant analysis (LDA) Multidimensional scaling (MDS) Non-negative matrix factorization (NMF) Partial least squares regression (PLSR) Principal component analysis (PCA) Principal component regression (PCR) Projection pursuit Sammon mapping t-distributed stochastic neighbor embedding (t-SNE) === Ensemble learning === Ensemble learning AdaBoost Boosting Bootstrap aggregating (also "bagging" or "bootstrapping") Ensemble averaging Gradient boosted decision tree (GBDT) Gradient boosting Random Forest Stacked Generalization === Meta-learning === Meta-learning Inductive bias Metadata === Reinforcement learning === Reinforcement learning Q-learning State–action–reward–state–action (SARSA) Temporal difference learning (TD) Learning Automata === Supervised learning === Supervised learning Averaged one-dependence estimators (AODE) Artificial neural network Case-based reasoning Gaussian process regression Gene expression programming Group method of data handling (GMDH) Inductive logic programming Instance-based learning Lazy learning Learning Automata Learning Vector Quantization Logistic Model Tree Minimum message length (decision trees, decision graphs, etc.) Nearest Neighbor Algorithm Analogical modeling Probably approximately correct learning (PAC) learning Ripple down rules, a knowledge acquisition methodology Symbolic machine learning algorithms Support vector machines Random Forests Ensembles of classifiers Bootstrap aggregating (bagging) Boosting (meta-algorithm) Ordinal classification Conditional Random Field ANOVA Quadratic classifiers k-nearest neighbor Boosting SPRINT Bayesian networks Naive Bayes Hidden Markov models Hierarchical hidden Markov model ==== Bayesian ==== Bayesian statistics Bayesian knowledge base Naive Bayes Gaussian Naive Bayes Multinomial Naive Bayes Averaged One-Dependence Estimators (AODE) Bayesian Belief Network (BBN) Bayesian Network (BN) ==== Decision tree algorithms ==== Decision tree algorithm Decision tree Classification and regression tree (CART) Iterative Dichotomiser 3 (ID3) C4.5 algorithm C5.0 algorithm Chi-squared Automatic Interaction Detection (CHAID) Decision stump Conditional decision tree ID3 algorithm Random forest SLIQ ==== Linear classifier ==== Linear classifier Fisher's linear discriminant Linear regression Logistic regression Multinomial logistic regression Naive Bayes classifier Perceptron Support vector machine === Unsupervised learning === Unsupervised learning Expectation-maximization algorithm Vector Quantization Generative topographic map Information bottleneck method Association rule learning algorithms Apriori algorithm Eclat algorithm ==== Artificial neural networks ==== Artificial neural network Feedforward neural network Extreme learning machine Convolutional neural network Recurrent neural network Long short-term memory (LSTM) Logic learning machine Self-organizing map ==== Association rule learning ==== Association rule learning Apriori algorithm Eclat algorithm FP-growth algorithm ==== Hierarchical clustering ==== Hierarchical clustering Single-linkage clustering Conceptual clustering ==== Cluster analysis ==== Cluster analysis BIRCH DBSCAN Expectation–maximization (EM) Fuzzy clustering Hierarchical clustering k-means clustering k-medians Mean-shift OPTICS algorithm ==== Anomaly detection ==== Anomaly detection k-nearest neighbors algorithm (k-NN) Local outlier factor === Semi-supervised learning === Semi-supervised learning Active learning Generative models Low-density separation Graph-based methods Co-training Transduction === Deep learning === Deep learning Deep belief networks Deep Boltzmann machines Deep Convolutional neural networks Deep Recurrent neural networks Hierarchical temporal memory Generative Adversarial Network Style transfer Transformer Stacked Auto-Encoders === Other machine learning methods and problems === Anomaly detection Association rules Bias-variance dilemma Classification Multi-label classification Clustering Data Pre-processing Empirical risk minimization Feature engineering Feature learning Learning to rank Occam learning Online machine learning PAC learning Regression Reinforcement Learning Semi-supervised learning Statistical learning Structured prediction Graphical models Bayesian network Conditional random field (CRF) Hidden Markov model (HMM) Unsupervised learning VC theory == Machine learning research == List of artificial intelligence projects List of datasets for machine learning research == History of machine learning == History of machine learning Timeline of machine learning == Machine learning projects == Machine learning projects: DeepMind Google Brain OpenAI Meta AI Hugging Face == Machine learning organizations == === Machine learning conferences and workshops === Artificial Intelligence and Security (AISec) (co-located workshop with CCS) Conference on Neural Information Processing Systems (NIPS) ECML PKDD International Conference on Machine Learning (ICML) ML4ALL (Machine Learning For All) == Machine learning publications == === Books on machine learning === Mathematics for Machine Learning Hands-On Machine Learning Scikit-Learn, Keras, and TensorFlow The Hundred-Page Machine Learning Book === Machine learning journals === Machine Learning Journal of Machine Learning Research (JMLR) Neural Computation == Pe
Read more →
Deterministic acyclic finite state automaton

In computer science, a deterministic acyclic finite state automaton (DAFSA), is a data structure that represents a set of strings, and allows for a query operation that tests whether a given string belongs to the set in time proportional to its length. Algorithms exist to construct and maintain such automata, while keeping them minimal. DAFSA is the rediscovery of a data structure called Directed Acyclic Word Graph (DAWG), although the same name had already been given to a different data structure which is related to suffix automaton. A DAFSA is a special case of a finite state recognizer that takes the form of a directed acyclic graph with a single source vertex (a vertex with no incoming edges), in which each edge of the graph is labeled by a letter or symbol, and in which each vertex has at most one outgoing edge for each possible letter or symbol. The strings represented by the DAFSA are formed by the symbols on paths in the graph from the source vertex to any sink vertex (a vertex with no outgoing edges). In fact, a deterministic finite state automaton is acyclic if and only if it recognizes a finite set of strings. == History == Blumer et al first defined terminology Directed Acyclic Word Graph (DAWG) in 1983. Appel and Jacobsen used the same naming for a different data structure in 1988. Independent of earlier work, Daciuk et al rediscovered the latter data structure in 2000 but called it DAFSA. == Comparison to tries == By allowing the same vertices to be reached by multiple paths, a DAFSA may use significantly fewer vertices than the strongly related trie data structure. Consider, for example, the four English words "tap", "taps", "top", and "tops". A trie for those four words would have 12 vertices, one for each of the strings formed as a prefix of one of these words, or for one of the words followed by the end-of-string marker. However, a DAFSA can represent these same four words using only six vertices vi for 0 ≤ i ≤ 5, and the following edges: an edge from v0 to v1 labeled "t", two edges from v1 to v2 labeled "a" and "o", an edge from v2 to v3 labeled "p", an edge v3 to v4 labeled "s", and edges from v3 and v4 to v5 labeled with the end-of-string marker. There is a tradeoff between memory and functionality, because a standard DAFSA can tell you if a word exists within it, but it cannot point you to auxiliary information about that word, whereas a trie can. The primary difference between DAFSA and trie is the elimination of suffix and infix redundancy in storing strings. The trie eliminates prefix redundancy since all common prefixes are shared between strings, such as between doctors and doctorate the doctor prefix is shared. In a DAFSA common suffixes are also shared, for words that have the same set of possible suffixes as each other. For dictionary sets of common English words, this translates into major memory usage reduction. Because the terminal nodes of a DAFSA can be reached by multiple paths, a DAFSA cannot directly store auxiliary information relating to each path, e.g. a word's frequency in the English language. However, if for each node we store the number of unique paths through that point in the structure, we can use it to retrieve the index of a word, or a word given its index. The auxiliary information can then be stored in an array.
Read more →
Ocrad

Ocrad is an optical character recognition program and part of the GNU Project. It is free software licensed under the GNU GPL. Based on a feature extraction method, it reads images in portable pixmap formats known as Portable anymap and produces text in byte (8-bit) or UTF-8 formats. Also included is a layout analyser, able to separate the columns or blocks of text normally found on printed pages. == User interface == Ocrad can be used as a stand-alone command-line application or as a back-end to other programs. Kooka, which was the KDE environment's default scanning application until KDE 4, can use Ocrad as its OCR engine. Since conversion to newer Qt versions, current versions of KDE no longer contain Kooka; development continues in the KDE git repository. Ocrad can be also used as an OCR engine in OCRFeeder. == History == Ocrad has been developed by Antonio Diaz Diaz since 2003. Version 0.7 was released in February 2004, 0.14 in February 2006 and 0.18 in May 2009. It is written in C++. Archives of the bug-ocrad mailing list go back to October 2003.
Read more →
Rada Mihalcea

Rada Mihalcea is the Janice M. Jenkins Collegiate Professor of Computer Science and Engineering at the University of Michigan. She has made significant contributions to natural language processing, multimodal processing, computational social science, and AI for Social Good. With Paul Tarau, she invented the TextRank Algorithm, which is a classic algorithm widely used for text summarization. == Career == Mihalcea has a Ph.D. in Computer Science and Engineering from Southern Methodist University (2001) and a Ph.D. in Linguistics, Oxford University (2010). In 2017 she was named Director of the Artificial Intelligence Laboratory at University of Michigan, Computer Science and Engineering. In 2018, Mihalcea was elected as vice president for the Association for Computational Linguistics (ACL). In 2021, she was elected the president for ACL. She is a professor of Computer Science and Engineering at the University of Michigan, where she also leads the Language and Information Technologies (LIT) Lab. Before joining UofM, she was a professor at North Texas University between 2002-2013. A prolific researcher, Mihalcea has authored or coauthored over 500 articles since 1998 on topics ranging from semantic analysis of text to lie detection. Her work has been cited over 50,000 times on Google Scholar, which made her one of the most cited scholars in Multimodal Interaction and Computational Social Science. In 2008, Mihalcea received the Presidential Early Career Award for Scientists and Engineers (PECASE) She is an ACM Fellow (since 2019), AAAI Fellow (since 2021), and ACL Fellow (since 2025). Mihalcea is an outspoken promoter of diversity in computer science. She also supports an expansion of the traditional analysis of educational success, which tends to focus on academic behaviour, to include student life, personality and background outside of the classroom. Mihalcea leads Girls Encoded, a program designed to develop the pipeline of women in computer science as well as to retain the women who have entered into the program. == Awards == Elected to American Academy of Arts & Sciences, 2026 ACL Fellow, 2025 "for significant contributions to graph-based language processing, computational social science, and the advancement of NLP for social good." AAAI Fellow, 2021 "for significant contributions to natural language processing and computational social science". ACM Fellow, 2019 "for contributions to natural language processing, with innovations in data-driven and graph-based language processing". Sarah Goddard Power Award, 2019. Carol Hollenshead Award, 2018. Presidential Early Career Award for Scientists and Engineers (PECASE), 2009. Awarded by President Barack Obama. == Research == Mihalcea is known for her research in natural language processing, multimodal processing, computational social sciences. In a collaboration she leads at the University of Michigan, Mihalcea has created software that can detect human lying. In a study of video clips of high profile court cases, a computer was more accurate at detecting deception than human judges. Mihalcea's lie-detection software uses machine learning techniques to analyze video clips of actual trials. In her 2015 study, the team used clips from The Innocence Project, a national organization that works to reexamine cases where individuals were tried without the benefit of DNA testing with the aim of exonerating wrongfully convicted individuals. After identifying common human gestures, they transcribed the audio from the video clips of trials and analyzed how often subjects labeled deceptive used various words and phrases. The system was 75% accurate in identifying which subjects were deceptive among 120 videos. That puts Mihalcea's algorithm on par with the most commonly accepted form of lie detection, polygraph tests, which are roughly 85 percent accurate when testing guilty people and 56 percent accurate when testing the innocent. She notes there are still improvements to be made — in particular to account for cultural and demographic differences. A possibly unique advantage of Mihalcea's study was the real world, high stakes nature of the footage analyzed in the study. In laboratory experiments, it is difficult to create a setting that motivates people to truly lie. In 2018, Mihalcea and her collaborators worked on an algorithm-based system that identifies linguistic cues in fake news stories. It successfully found fakes up to 76% of the time, compared to a human success rate of 70%. == Publications == === Books === Rada Mihalcea and Dragomir Radev, Graph-based Natural Language Processing and Information Retrieval, Cambridge U. Press, 2011. Gabe Ignatow and Rada Mihalcea, Text Mining: A Guidebook for the Social Sciences, SAGE, 2016. Gabe Ignatow and Rada Mihalcea, An Introduction to Text Mining: Research Design, Data Collection, and Analysis, SAGE, 2017. === Journals and conferences === Textrank: Bringing order into text. R. Mihalcea, P. Tarau. Proceedings of the 2004 conference on empirical methods in natural language processing. 2004 Corpus-based and knowledge-based measures of text semantic similarity. R. Mihalcea, C. Corley, C. Strapparava. AAAI 6, 775-780. 2006 Wikify!: linking documents to encyclopedic knowledge. R. Mihalcea, A. Csomai. Proceedings of the sixteenth ACM conference on Conference on information and information management. 2007 Learning to identify emotions in text. C. Strapparava, R. Mihalcea. Proceedings of the 2008 ACM symposium on Applied computing, 1556-1560. 2008 Semeval-2007 task 14: Affective text. C. Strapparava, R. Mihalcea. Proceedings of the Fourth International Workshop on Semantic Evaluations. 2007 Learning multilingual subjective language via cross-lingual projections. R. Mihalcea, C. Banea, J. Wiebe. Proceedings of the 45th annual meeting of the association of computational linguistics. 2007 Graph-based ranking algorithms for sentence extraction, applied to text summarization. R. Mihalcea. Proceedings of the ACL Interactive Poster and Demonstration Sessions. 2004 Falcon: Boosting knowledge for answer engines. S. Harabagiu, D. Moldovan, M. Pasca, R. Mihalcea, M. Surdeanu, Razvan Bunescu, Roxana Girju, Vasile Rus, Paul Morarescu. TREC 9, 479-488. 2000 Measuring the semantic similarity of texts. C. Corley, R. Mihalcea. Proceedings of the ACL workshop on empirical modeling of semantic equivalence and entailment. 2005 R Mihalcea (2007). "Using wikipedia for automatic word-sense disambiguation". Human Language Technologies 2007: The Conference of the North American Chapter of the Association for Computational Linguistics; Proceedings of the Main Conference. CiteSeerX 10.1.1.74.3561. - see also Word-sense disambiguation Unsupervised graph-based word sense disambiguation using measures of word semantic similarity. R. Sinha, R. Mihalcea. International Conference on Semantic Computing (ICSC 2007), 363-369. 2007 == Personal life == Mihalcea was born in Cluj-Napoca, Romania, where she attended the Technical University of Cluj-Napoca. She can speak Romanian, English, Italian, and French. Mihalcea has two children - Zara (b. 2009) and Caius (b. 2013). They were both born in Dallas, Texas. She is married to an associate professor of engineering at the University of Michigan–Flint - Mihai Burzo. They met while they were both completing Ph.D.s at Southern Methodist University in 2001 and have often collaborated on research, such as the 2015 study on lie detection.
Read more →
Color picker

A color picker (also color chooser or color tool) is a graphical user interface widget, usually found within graphics software or online, used to select colors and, in some cases, to create color schemes (the color picker might be more sophisticated than the palette included with the program). Operating systems such as Microsoft Windows or macOS have a system color picker, which can be used by third-party programs (e.g., Adobe Photoshop). == History == The concept of color pickers dates back to the early days of computer graphics and digital design. Early versions were rudimentary, often featuring basic color palettes and limited functionality. One of the first drawing programs to include a color picker was SketchPad (also referred to as LisaSketch), designed by Bill Atkinson in 1983 to showcase LisaGraf's capabilities. It used a black and white pattern system, using dithering to create the illusion of color depth. With the increased popularity of personal computers with color graphics, there soon came software similar to SketchPad that supported more than two colors, like Broderbund's Dazzle Draw for the Apple II or Electronic Arts' Deluxe Paint. However, the color pickers present in those programs relied on indexed colors. Color pickers, resembling ones used in modern software with support for direct, 24-bit color, appeared soon after the release of the Macintosh II, with the release of programs like Adobe Photoshop and Corel Painter. As the increase of color depth allowed the choice of significantly more colors, the shape and form of color pickers started to diverge. For example, Adobe Photoshop used a hue-saturation color wheel with a slider for brightness in version 0.63, later on switching to a rectangular design accompanied by a hue slider. Corel Painter pioneered the triangular saturation and brightness picker with a hue ring around it, aiming to better represent the continuity of the hue spectrum and the relationship between saturation and brightness. == Purpose == A color picker is used to select and adjust color values. In graphic design and image editing, users typically choose colors via an interface with a visual representation of a color—organized with quasi-perceptually-relevant hue, saturation and lightness dimensions (HSL) – instead of keying in alphanumeric text values. Because color appearance depends on comparison of neighboring colors (see color vision), many interfaces attempt to clarify the relationships between colors. == Interface == Color tools can vary in their interface. Some may use sliders, buttons, text boxes for color values, or direct manipulation. Often a two-dimensional square is used to create a range of color values (such as lightness and saturation) that can be clicked on or selected in some other manner. Drag and drop, color droppers, and various other forms of interfaces are commonly used as well. Usually, color values are also displayed numerically, so they can be precisely remembered and keyed-in later, such as three values of 0-255 representing red, green, and blue, respectively. === Eyedropper === The eyedropper is a tool present in most color pickers and graphics software that allows a user to read a color at a specific point in an image, or position on a display. This enables the color to be transferred to other applications particularly quickly. Modern implementations of eyedropper tools are also available as browser extensions, allowing users to pick colors directly from web pages, such as in Google Chrome and Microsoft Edge. == Working == A color picker has two main parts, first a color slider and second a color canvas. The color slider has a linear or radial gradient of the seven rainbow colors i.e. Violet, Indigo, Blue, Green, Yellow, Orange and Red. It allows one to choose any of the seven primary colors. The color value chosen from the color slider instantly reflects in the color canvas. The color canvas is a mixture of two linear color gradients. First a linear gradient of the current chosen color and second a linear gradient of the black color. This mixture of color gradients lets one choose a lighter and darker version of the current chosen color from the color slider.
Read more →
The Best Free AI Sales Assistant for Beginners

Comparing the best AI sales assistant? An AI sales assistant is software that uses machine learning to help you get more done — it lowers the barrier so anyone can produce professional output. Privacy matters too: check whether your data trains the model and whether a no-log or enterprise tier is available. Whether you are a beginner or a pro, the right AI sales assistant slots into your workflow and pays for itself fast. We tested the leading options and ranked them by quality, value, and ease of use.
Read more →
Best AI Sales Assistants in 2026

Shopping for the best AI sales assistant? An AI sales assistant 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 sales assistant slots into your workflow and pays for itself fast. We tested the leading options and ranked them by quality, value, and ease of use.
Read more →
Google matrix

A Google matrix is a particular stochastic matrix that is used by Google's PageRank algorithm. The matrix represents a graph with edges representing links between pages. The PageRank of each page can then be generated iteratively from the Google matrix using the power method. However, in order for the power method to converge, the matrix must be stochastic, irreducible and aperiodic. == Adjacency matrix A and Markov matrix S == In order to generate the Google matrix G, we must first generate an adjacency matrix A which represents the relations between pages or nodes. Assuming there are N pages, we can fill out A by doing the following: A matrix element A i , j {\displaystyle A_{i,j}} is filled with 1 if node j {\displaystyle j} has a link to node i {\displaystyle i} , and 0 otherwise; this is the adjacency matrix of links. A related matrix S corresponding to the transitions in a Markov chain of given network is constructed from A by dividing the elements of column "j" by a number of k j = Σ i = 1 N A i , j {\displaystyle k_{j}=\Sigma _{i=1}^{N}A_{i,j}} where k j {\displaystyle k_{j}} is the total number of outgoing links from node j to all other nodes. The columns having zero matrix elements, corresponding to dangling nodes, are replaced by a constant value 1/N. Such a procedure adds a link from every sink, dangling state a {\displaystyle a} to every other node. Now by the construction the sum of all elements in any column of matrix S is equal to unity. In this way the matrix S is mathematically well defined and it belongs to the class of Markov chains and the class of Perron-Frobenius operators. That makes S suitable for the PageRank algorithm. == Construction of Google matrix G == Then the final Google matrix G can be expressed via S as: G i j = α S i j + ( 1 − α ) 1 N ( 1 ) {\displaystyle G_{ij}=\alpha S_{ij}+(1-\alpha ){\frac {1}{N}}\;\;\;\;\;\;\;\;\;\;\;(1)} By the construction the sum of all non-negative elements inside each matrix column is equal to unity. The numerical coefficient α {\displaystyle \alpha } is known as a damping factor. Usually S is a sparse matrix and for modern directed networks it has only about ten nonzero elements in a line or column, thus only about 10N multiplications are needed to multiply a vector by matrix G. == Examples of Google matrix == An example of the matrix S {\displaystyle S} construction via Eq.(1) within a simple network is given in the article CheiRank. For the actual matrix, Google uses a damping factor α {\displaystyle \alpha } around 0.85. The term ( 1 − α ) {\displaystyle (1-\alpha )} gives a surfer probability to jump randomly on any page. The matrix G {\displaystyle G} belongs to the class of Perron-Frobenius operators of Markov chains. The examples of Google matrix structure are shown in Fig.1 for Wikipedia articles hyperlink network in 2009 at small scale and in Fig.2 for University of Cambridge network in 2006 at large scale. == Spectrum and eigenstates of G matrix == For 0 < α < 1 {\displaystyle 0<\alpha <1} there is only one maximal eigenvalue λ = 1 {\displaystyle \lambda =1} with the corresponding right eigenvector which has non-negative elements P i {\displaystyle P_{i}} which can be viewed as stationary probability distribution. These probabilities ordered by their decreasing values give the PageRank vector P i {\displaystyle P_{i}} with the PageRank K i {\displaystyle K_{i}} used by Google search to rank webpages. Usually one has for the World Wide Web that P ∝ 1 / K β {\displaystyle P\propto 1/K^{\beta }} with β ≈ 0.9 {\displaystyle \beta \approx 0.9} . The number of nodes with a given PageRank value scales as N P ∝ 1 / P ν {\displaystyle N_{P}\propto 1/P^{\nu }} with the exponent ν = 1 + 1 / β ≈ 2.1 {\displaystyle \nu =1+1/\beta \approx 2.1} . The left eigenvector at λ = 1 {\displaystyle \lambda =1} has constant matrix elements. With 0 < α {\displaystyle 0<\alpha } all eigenvalues move as λ i → α λ i {\displaystyle \lambda _{i}\rightarrow \alpha \lambda _{i}} except the maximal eigenvalue λ = 1 {\displaystyle \lambda =1} , which remains unchanged. The PageRank vector varies with α {\displaystyle \alpha } but other eigenvectors with λ i < 1 {\displaystyle \lambda _{i}<1} remain unchanged due to their orthogonality to the constant left vector at λ = 1 {\displaystyle \lambda =1} . The gap between λ = 1 {\displaystyle \lambda =1} and other eigenvalue being 1 − α ≈ 0.15 {\displaystyle 1-\alpha \approx 0.15} gives a rapid convergence of a random initial vector to the PageRank approximately after 50 multiplications on G {\displaystyle G} matrix. At α = 1 {\displaystyle \alpha =1} the matrix G {\displaystyle G} has generally many degenerate eigenvalues λ = 1 {\displaystyle \lambda =1} (see e.g. [6]). Examples of the eigenvalue spectrum of the Google matrix of various directed networks is shown in Fig.3 from and Fig.4 from. The Google matrix can be also constructed for the Ulam networks generated by the Ulam method [8] for dynamical maps. The spectral properties of such matrices are discussed in [9,10,11,12,13,15]. In a number of cases the spectrum is described by the fractal Weyl law [10,12]. The Google matrix can be constructed also for other directed networks, e.g. for the procedure call network of the Linux Kernel software introduced in [15]. In this case the spectrum of λ {\displaystyle \lambda } is described by the fractal Weyl law with the fractal dimension d ≈ 1.3 {\displaystyle d\approx 1.3} (see Fig.5 from ). Numerical analysis shows that the eigenstates of matrix G {\displaystyle G} are localized (see Fig.6 from ). Arnoldi iteration method allows to compute many eigenvalues and eigenvectors for matrices of rather large size [13]. Other examples of G {\displaystyle G} matrix include the Google matrix of brain [17] and business process management [18], see also. Applications of Google matrix analysis to DNA sequences is described in [20]. Such a Google matrix approach allows also to analyze entanglement of cultures via ranking of multilingual Wikipedia articles abouts persons [21] == Historical notes == The Google matrix with damping factor was described by Sergey Brin and Larry Page in 1998 [22], see also articles on PageRank history [23], [24].
Read more →
Sample complexity

The sample complexity of a machine learning algorithm represents the number of training-samples that it needs in order to successfully learn a target function. More precisely, the sample complexity is the number of training-samples that we need to supply to the algorithm, so that the function returned by the algorithm is within an arbitrarily small error of the best possible function, with probability arbitrarily close to 1. There are two variants of sample complexity: The weak variant fixes a particular input-output distribution; The strong variant takes the worst-case sample complexity over all input-output distributions. The No free lunch theorem, discussed below, proves that, in general, the strong sample complexity is infinite, i.e. that there is no algorithm that can learn the globally-optimal target function using a finite number of training samples. However, if we are only interested in a particular class of target functions (e.g., only linear functions) then the sample complexity is finite, and it depends linearly on the VC dimension on the class of target functions. == Definition == Let X {\displaystyle X} be a space which we call the input space, and Y {\displaystyle Y} be a space which we call the output space, and let Z {\displaystyle Z} denote the product X × Y {\displaystyle X\times Y} . For example, in the setting of binary classification, X {\displaystyle X} is typically a finite-dimensional vector space and Y {\displaystyle Y} is the set { − 1 , 1 } {\displaystyle \{-1,1\}} . Fix a hypothesis space H {\displaystyle {\mathcal {H}}} of functions h : X → Y {\displaystyle h\colon X\to Y} . A learning algorithm over H {\displaystyle {\mathcal {H}}} is a computable map from Z {\displaystyle Z} to H {\displaystyle {\mathcal {H}}} . In other words, it is an algorithm that takes as input a finite sequence of training samples and outputs a function from X {\displaystyle X} to Y {\displaystyle Y} . Typical learning algorithms include empirical risk minimization, without or with Tikhonov regularization. Fix a loss function L : Y × Y → R ≥ 0 {\displaystyle {\mathcal {L}}\colon Y\times Y\to \mathbb {R} _{\geq 0}} , for example, the square loss L ( y , y ′ ) = ( y − y ′ ) 2 {\displaystyle {\mathcal {L}}(y,y')=(y-y')^{2}} , where h ( x ) = y ′ {\displaystyle h(x)=y'} . For a given distribution ρ {\displaystyle \rho } on X × Y {\displaystyle X\times Y} , the expected risk of a hypothesis (a function) h ∈ H {\displaystyle h\in {\mathcal {H}}} is E ( h ) := E ρ [ L ( h ( x ) , y ) ] = ∫ X × Y L ( h ( x ) , y ) d ρ ( x , y ) {\displaystyle {\mathcal {E}}(h):=\mathbb {E} _{\rho }[{\mathcal {L}}(h(x),y)]=\int _{X\times Y}{\mathcal {L}}(h(x),y)\,d\rho (x,y)} In our setting, we have h = A ( S n ) {\displaystyle h={\mathcal {A}}(S_{n})} , where A {\displaystyle {\mathcal {A}}} is a learning algorithm and S n = ( ( x 1 , y 1 ) , … , ( x n , y n ) ) ∼ ρ n {\displaystyle S_{n}=((x_{1},y_{1}),\ldots ,(x_{n},y_{n}))\sim \rho ^{n}} is a sequence of vectors which are all drawn independently from ρ {\displaystyle \rho } . Define the optimal risk E H ∗ = inf h ∈ H E ( h ) . {\displaystyle {\mathcal {E}}_{\mathcal {H}}^{}={\underset {h\in {\mathcal {H}}}{\inf }}{\mathcal {E}}(h).} Set h n = A ( S n ) {\displaystyle h_{n}={\mathcal {A}}(S_{n})} , for each sample size n {\displaystyle n} . h n {\displaystyle h_{n}} is a random variable and depends on the random variable S n {\displaystyle S_{n}} , which is drawn from the distribution ρ n {\displaystyle \rho ^{n}} . The algorithm A {\displaystyle {\mathcal {A}}} is called consistent if E ( h n ) {\displaystyle {\mathcal {E}}(h_{n})} probabilistically converges to E H ∗ {\displaystyle {\mathcal {E}}_{\mathcal {H}}^{}} . In other words, for all ϵ , δ > 0 {\displaystyle \epsilon ,\delta >0} , there exists a positive integer N {\displaystyle N} , such that, for all sample sizes n ≥ N {\displaystyle n\geq N} , we have Pr ρ n [ E ( h n ) − E H ∗ ≥ ε ] < δ . {\displaystyle \Pr _{\rho ^{n}}[{\mathcal {E}}(h_{n})-{\mathcal {E}}_{\mathcal {H}}^{}\geq \varepsilon ]<\delta .} The sample complexity of A {\displaystyle {\mathcal {A}}} is then the minimum N {\displaystyle N} for which this holds, as a function of ρ , ϵ {\displaystyle \rho ,\epsilon } , and δ {\displaystyle \delta } . We write the sample complexity as N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} to emphasize that this value of N {\displaystyle N} depends on ρ , ϵ {\displaystyle \rho ,\epsilon } , and δ {\displaystyle \delta } . If A {\displaystyle {\mathcal {A}}} is not consistent, then we set N ( ρ , ϵ , δ ) = ∞ {\displaystyle N(\rho ,\epsilon ,\delta )=\infty } . If there exists an algorithm for which N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} is finite, then we say that the hypothesis space H {\displaystyle {\mathcal {H}}} is learnable. In others words, the sample complexity N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} defines the rate of consistency of the algorithm: given a desired accuracy ϵ {\displaystyle \epsilon } and confidence δ {\displaystyle \delta } , one needs to sample N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} data points to guarantee that the risk of the output function is within ϵ {\displaystyle \epsilon } of the best possible, with probability at least 1 − δ {\displaystyle 1-\delta } . In probably approximately correct (PAC) learning, one is concerned with whether the sample complexity is polynomial, that is, whether N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} is bounded by a polynomial in 1 / ϵ {\displaystyle 1/\epsilon } and 1 / δ {\displaystyle 1/\delta } . If N ( ρ , ϵ , δ ) {\displaystyle N(\rho ,\epsilon ,\delta )} is polynomial for some learning algorithm, then one says that the hypothesis space H {\displaystyle {\mathcal {H}}} is PAC-learnable. This is a stronger notion than being learnable. == Unrestricted hypothesis space: infinite sample complexity == One can ask whether there exists a learning algorithm so that the sample complexity is finite in the strong sense, that is, there is a bound on the number of samples needed so that the algorithm can learn any distribution over the input-output space with a specified target error. More formally, one asks whether there exists a learning algorithm A {\displaystyle {\mathcal {A}}} , such that, for all ϵ , δ > 0 {\displaystyle \epsilon ,\delta >0} , there exists a positive integer N {\displaystyle N} such that for all n ≥ N {\displaystyle n\geq N} , we have sup ρ ( Pr ρ n [ E ( h n ) − E H ∗ ≥ ε ] ) < δ , {\displaystyle \sup _{\rho }\left(\Pr _{\rho ^{n}}[{\mathcal {E}}(h_{n})-{\mathcal {E}}_{\mathcal {H}}^{}\geq \varepsilon ]\right)<\delta ,} where h n = A ( S n ) {\displaystyle h_{n}={\mathcal {A}}(S_{n})} , with S n = ( ( x 1 , y 1 ) , … , ( x n , y n ) ) ∼ ρ n {\displaystyle S_{n}=((x_{1},y_{1}),\ldots ,(x_{n},y_{n}))\sim \rho ^{n}} as above. The No Free Lunch Theorem says that without restrictions on the hypothesis space H {\displaystyle {\mathcal {H}}} , this is not the case, i.e., there always exist "bad" distributions for which the sample complexity is arbitrarily large. Thus, in order to make statements about the rate of convergence of the quantity sup ρ ( Pr ρ n [ E ( h n ) − E H ∗ ≥ ε ] ) , {\displaystyle \sup _{\rho }\left(\Pr _{\rho ^{n}}[{\mathcal {E}}(h_{n})-{\mathcal {E}}_{\mathcal {H}}^{}\geq \varepsilon ]\right),} one must either constrain the space of probability distributions ρ {\displaystyle \rho } , e.g. via a parametric approach, or constrain the space of hypotheses H {\displaystyle {\mathcal {H}}} , as in distribution-free approaches. == Restricted hypothesis space: finite sample-complexity == The latter approach leads to concepts such as VC dimension and Rademacher complexity which control the complexity of the space H {\displaystyle {\mathcal {H}}} . A smaller hypothesis space introduces more bias into the inference process, meaning that E H ∗ {\displaystyle {\mathcal {E}}_{\mathcal {H}}^{}} may be greater than the best possible risk in a larger space. However, by restricting the complexity of the hypothesis space it becomes possible for an algorithm to produce more uniformly consistent functions. This trade-off leads to the concept of regularization. It is a theorem from VC theory that the following three statements are equivalent for a hypothesis space H {\displaystyle {\mathcal {H}}} : H {\displaystyle {\mathcal {H}}} is PAC-learnable. The VC dimension of H {\displaystyle {\mathcal {H}}} is finite. H {\displaystyle {\mathcal {H}}} is a uniform Glivenko-Cantelli class. This gives a way to prove that certain hypothesis spaces are PAC learnable, and by extension, learnable. === An example of a PAC-learnable hypothesis space === X = R d , Y = { − 1 , 1 } {\displaystyle X=\mathbb {R} ^{d},Y=\{-1,1\}} , and let H {\displaystyle {\mathcal {H}}} be the space of affine functions on X {\displaystyle X} , that is, functions of the form x ↦ ⟨ w , x ⟩ + b {\displaystyle x\mapsto \langl
Read more →
Wang-Chiew Tan

Wang-Chiew Tan is a Singaporean computer scientist specializing in data management and natural language processing. Her work in data management includes data provenance (or data lineage) and data integration. She is currently a Research Scientist at Facebook AI, and was previously the Director of Research at Megagon Labs in Mountain View, California. At Megagon Labs, Tan was the lead researcher on a study with the University of Tokyo that concluded that the company of other people is more effective than pets at making people happy. == Education and career == Tan earned her bachelor's degree in computer science (first-class) at the National University of Singapore, and completed her Ph.D. at the University of Pennsylvania. Her 2002 dissertation, Data Annotations, Provenance, and Archiving, was jointly supervised by Peter Buneman and Sanjeev Khanna. Before working at Megagon, she has been a professor of computer science at the University of California, Santa Cruz beginning in 2002, and, from 2010 to 2012, was on leave from Santa Cruz as a researcher at IBM Research - Almaden. == Recognition == Tan was named a Fellow of the Association for Computing Machinery in 2015 "for contributions to data provenance and to the foundations of information integration".
Read more →
Michael Collins (computational linguist)

Michael J. Collins (born 4 March 1970) is a researcher in the field of computational linguistics. He is the Vikram S. Pandit Professor of Computer Science at Columbia University. His research interests are in natural language processing as well as machine learning and he has made important contributions in statistical parsing and in statistical machine learning. In his studies Collins covers a wide range of topics such as parse re-ranking, tree kernels, semi-supervised learning, machine translation and exponentiated gradient algorithms with a general focus on discriminative models and structured prediction. One notable contribution is a state-of-the-art parser for the Penn Wall Street Journal corpus. As of 11 November 2015, his works have been cited 16,020 times, and he has an h-index of 47. Collins worked as a researcher at AT&T Labs between January 1999 and November 2002, and later held the positions of assistant and associate professor at M.I.T. Since January 2011, he has been a professor at Columbia University. In 2011, he was named a fellow of the Association for Computational Linguistics.
Read more →
AI Marketing Tools: Free vs Paid (2026)

Shopping for the best AI marketing tool? An AI marketing tool 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 marketing tool 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 →
Graph cut optimization

Graph cut optimization is a combinatorial optimization method applicable to a family of functions of discrete variables, named after the concept of cut in the theory of flow networks. Thanks to the max-flow min-cut theorem, determining the minimum cut over a graph representing a flow network is equivalent to computing the maximum flow over the network. Given a pseudo-Boolean function f {\displaystyle f} , if it is possible to construct a flow network with positive weights such that each cut C {\displaystyle C} of the network can be mapped to an assignment of variables x {\displaystyle \mathbf {x} } to f {\displaystyle f} (and vice versa), and the cost of C {\displaystyle C} equals f ( x ) {\displaystyle f(\mathbf {x} )} (up to an additive constant) then it is possible to find the global optimum of f {\displaystyle f} in polynomial time by computing a minimum cut of the graph. The mapping between cuts and variable assignments is done by representing each variable with one node in the graph and, given a cut, each variable will have a value of 0 if the corresponding node belongs to the component connected to the source, or 1 if it belong to the component connected to the sink. Not all pseudo-Boolean functions can be represented by a flow network, and in the general case the global optimization problem is NP-hard. There exist sufficient conditions to characterise families of functions that can be optimised through graph cuts, such as submodular quadratic functions. Graph cut optimization can be extended to functions of discrete variables with a finite number of values, that can be approached with iterative algorithms with strong optimality properties, computing one graph cut at each iteration. Graph cut optimization is an important tool for inference over graphical models such as Markov random fields or conditional random fields, and it has applications in computer vision problems such as image segmentation, denoising, registration and stereo matching. == Representability == A pseudo-Boolean function f : { 0 , 1 } n → R {\displaystyle f:\{0,1\}^{n}\to \mathbb {R} } is said to be representable if there exists a graph G = ( V , E ) {\displaystyle G=(V,E)} with non-negative weights and with source and sink nodes s {\displaystyle s} and t {\displaystyle t} respectively, and there exists a set of nodes V 0 = { v 1 , … , v n } ⊂ V − { s , t } {\displaystyle V_{0}=\{v_{1},\dots ,v_{n}\}\subset V-\{s,t\}} such that, for each tuple of values ( x 1 , … , x n ) ∈ { 0 , 1 } n {\displaystyle (x_{1},\dots ,x_{n})\in \{0,1\}^{n}} assigned to the variables, f ( x 1 , … , x n ) {\displaystyle f(x_{1},\dots ,x_{n})} equals (up to a constant) the value of the flow determined by a minimum cut C = ( S , T ) {\displaystyle C=(S,T)} of the graph G {\displaystyle G} such that v i ∈ S {\displaystyle v_{i}\in S} if x i = 0 {\displaystyle x_{i}=0} and v i ∈ T {\displaystyle v_{i}\in T} if x i = 1 {\displaystyle x_{i}=1} . It is possible to classify pseudo-Boolean functions according to their order, determined by the maximum number of variables contributing to each single term. All first order functions, where each term depends upon at most one variable, are always representable. Quadratic functions f ( x ) = w 0 + ∑ i w i ( x i ) + ∑ i < j w i j ( x i , x j ) . {\displaystyle f(\mathbf {x} )=w_{0}+\sum _{i}w_{i}(x_{i})+\sum _{i 0 {\displaystyle p>0} then w i j k ( x i , x j , x k ) = w i j k ( 0 , 0 , 0 ) + p 1 ( x i − 1 ) + p 2 ( x j − 1 ) + p 3 ( x k − 1 ) + p 23 ( x j − 1 ) x k + p 31 x i ( x k − 1 ) + p 12 ( x i − 1 ) x j − p x i x j x k {\displaystyle w_{ijk}(x_{i},x_{j},x_{k})=w_{ijk}(0,0,0)+p_{1}(x_{i}-1)+p_{2}(x_{j}-1)+p_{3}(x_{k}-1)+p_{23}(x_{j}-1)x_{k}+p_{31}x_{i}(x_{k}-1)+p_{12}(x_{i}-1)x_{j}-px_{i}x_{j}x_{k}} with p 1 = w i j k ( 1 , 0 , 1 ) − w i j k ( 0 , 0 , 1 ) p 2 = w i j k ( 1 , 1 , 0 ) − w i j k ( 1 , 0 , 1 ) p 3 = w i j k ( 0 , 1 , 1 ) − w i j k ( 0 , 1 , 0 ) p 23 = w i j k ( 0 , 0 , 1 ) + w i j k ( 0 , 1 , 0 ) − w i j k ( 0 , 0 , 0 ) − w i j k ( 0 , 1 , 1 ) p 31 = w i j k ( 0 , 0 , 1 ) + w i j k ( 1 , 0 , 0 ) − w i j k ( 0 , 0 , 0 ) − w i j k ( 1 , 0 , 1 ) p 12 = w i j k ( 0 , 1 , 0 ) + w i j k ( 1 , 0 , 0 ) − w i j k ( 0 , 0 , 0 ) − w i j k ( 1 , 1 , 0 ) . {\displaystyle {\begin{aligned}p_{1}&=w_{ijk}(1,0,1)-w_{ijk}(0,0,1)\\p_{2}&=w_{ijk}(1,1,0)-w_{ijk}(1,0,1)\\p_{3}&=w_{ijk}(0,1,1)-w_{ijk}(0,1,0)\\p_{23}&=w_{ijk}(0,0,1)+w_{ijk}(0,1,0)-w_{ijk}(0,0,0)-w_{ijk}(0,1,1)\\p_{31}&=w_{ijk}(0,0,1)+w_{ijk}(1,0,0)-w_{ijk}(0,0,0)-w_{ijk}(1,0,1)\\p_{12}&=w_{ijk}(0,1,0)+w_{ijk}(1,0,0)-w_{ijk}(0,0,0)-w_{ijk}(1,1
Read more →
How to Choose an AI Copywriting Tool

Trying to pick the best AI copywriting tool? An AI copywriting tool is software that uses machine learning to help you get more done — it scales effortlessly from a single task to thousands. The best picks balance beginner-friendly simplicity with the depth power users need, and they ship updates often. Whether you are a beginner or a pro, the right AI copywriting tool 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 →
Project Bergamot

Project Bergamot is a joint project between several European universities and Mozilla for the development of machine translation software based on artificial neural networks, which is intended for local execution on end-user devices. The software library that was created and the associated language models were made available to the general public as Free Software. Execution requires a x86 CPU with SSE4.1 instruction set extensions. In 2022, Devin Coldewey of TechCrunch judged the translation quality to be "more than adequate", but considered Firefox Translations to be not yet fully mature. == Usage == Mozilla used the Bergamot Translator to expand its web browser Firefox with a feature for translating web pages, which was previously considered an important gap in Firefox' feature set. It is often compared to the much older corresponding feature in Google Chrome, which utilizes a cloud-based background service. In contrast, Firefox Translations does not require any data to leave the user's computer, resulting in advantages in terms of data protection, availability and possibly response times. There is just the installation of a new language model that needs to take place the first time a new language is encountered. Greater independence from large technology companies and their interests is also mentioned as an important advantage. Mozilla thus strengthened its position as an alternative software vendor with a particular focus on data protection and security. Mozilla followed up with the similar feature of speech recognition for spoken user input, based on whisperfile. On the other hand, slow translation times have been observed, especially on older devices. Also, Firefox Translations initially supported far fewer language pairs than other major translation services and is only gradually adding new models. On that matter, the training pipeline is also made available to interested parties to enable the creation of missing language models. TranslateLocally is a Firefox-independent translation software based on the Bergamot Translator. It is also available as an (Electron-based) standalone application or as an extension for Chromium-based web browsers. == History == Mozilla had already tried to get a (cloud-based) web content translation feature into Firefox a few years before Project Bergamot, but had failed because of the financial challenge. Microsoft had already delivered offline capabilities for its translation software in 2018. Google soon followed suit, Apple two years later. The software is based on the free translation framework Marian, which the University of Edinburgh had previously developed in cooperation with Microsoft, and is itself based on the Nematus toolkit that was presented in 2017. Under the leadership of the University of Edinburgh, a development consortium was formed with the Mozilla Corporation and the additional European universities of Prague, Sheffield and Tartu. In 2018, it was able to get 3 million euros of funding from the EU's Horizon 2020 programme. Firefox Translations was initially provided as an add-on. A first functional demonstration prototype was presented in October 2019. Beta version 117 had the feature integrated directly into the browser, the official release was in version 118 from September 2023. Both the add-on module and as part of Firefox, the code and the models are subject to the version 2 of the Mozilla Public License. Since 2022, the EU-funded HPLT project creates new language models. It involves additional partners, including the universities of Helsinki, Turku, Oslo and other partners from Spain, Norway and the Czech Republic.
Read more →