Machine learning made in a minute

The Accord.NET Framework is a .NET machine learning framework combined with audio and image processing libraries completely written in C#. It is a complete framework for building production-grade computer vision, computer audition, signal processing and statistics applications even for commercial use. A comprehensive set of sample applications provide a fast start to get up and running quickly, and an extensive documentation and wiki helps fill in the details.

Classification.

Support Vector Machines , Logistic Regression , Decision Trees , Neural Networks , Deep Learning (Deep Neural Networks) , Levenberg-Marquardt with Bayesian Regularization , Restricted Boltzmann Machines , Sequence classification , Hidden Markov Classifiers and Hidden Conditional Random Fields .

Regression.

Multiple linear regression , Multivariate linear regression , polynomial regression , logarithmic regression. Logistic regression , multinomial logistic regression (softmax) and generalized linear models . L2-regularized L2-loss logistic regression , L2-regularized logistic regression , L1-regularized logistic regression , L2-regularized logistic regression in the dual form and regression support vector machines .

Clustering.

K-Means , K-Modes , Mean-Shift , Gaussian Mixture Models , Binary Split , Deep Belief Networks , Restricted Boltzmann Machines . Clustering algorithms can be applied in arbitrary data, including images , data tables, videos and audio .

Distributions.

Parametric and non-parametric estimation of more than 40 distributions. Univariate distributions such as Normal , Cauchy , Hypergeometric , Poisson , Bernoulli , and specialized distributions such as the Kolmogorov-Smirnov , Nakagami , Weibull , and Von-Mises distributions. Multivariate distributions such as the multivariate Normal , Multinomial , Independent , Joint and Mixture distributions .

Hypothesis Tests.

More than 35 statistical hypothesis tests , including one way and two-way ANOVA tests , non-parametric tests such as the Kolmogorov-Smirnov test and the Sign Test for the Median , contingency table tests such as the Kappa test , with variations for multiple tables , as well as the Bhapkar and Bowker tests; and the more traditional Chi-Square , Z , F , T and Wald tests .

Kernel Methods.

Kernel Support Vector Machines , Multi-class and Multi-label machines , Sequential Minimal Optimization , Least-Squares Learning , probabilistic learning , including special methods for linear machines such as LIBLINEAR's methods for Linear Coordinate Descent , Linear Newton Method , Probabilistic Coordinate Descent , Probabilistic Coordinate Descent in the Dual , Probabilistic Newton Method for L1 and L2 machines in both the dual and primal formulations .

Imaging.

Interest and feature point detectors such as Harris , FREAK , SURF , and FAST . Grey-level Co-occurrence matrices , Border following , Bag-of-Visual-Words (BoW) , RANSAC-based homography estimation , integral images , haralick textural feature extraction , and dense descriptors such as histogram of oriented gradients (HOG) and Local Binary Pattern (LBP) . Several image filters for image processing applications such as difference of Gaussians , Gabor , Niblack and Sauvola thresholding .

Audio and Signal.

Load, parse, save, filter and transform audio signals , such as applying audio processing filters in both space and frequency domain . WAV files , audio capture , time-domain filters such as envelope , high-pass , low-pass , wave rectification filters. Frequency-domain operators such as differential rectification filter and comb filter with Dirac's delta functions . Signal generators for Cosine , Impulse , Square signals.

Vision.

Real-time face detection and tracking , as well as general methods for detecting , tracking and transforming objects in image streams . Contains cascade definitions , Camshift and Dynamic Template Matching trackers . Includes pre-created classifiers for human faces and some facial features such as noses .

Get started now!

Sample applications.

Sample applications help you start writing your applications quickly. Just get one of the sample applications that is closest to your goal , and start from there.

Create and test new learning algorithms with ease.

Strategy and template method patterns help you swap learning algorithms quickly. Create, build and compare different approaches without delving too deep in code. Check a list of works that have been made possible with the framework .

Ask your questions.

Stackoverflow is continuously monitored for new questions containing the "Accord.NET" tag. Ask your new question and mark it with this tag, and it will be answered in a minute by the framework authors and the user community.

Note: If you would like to ask questions concerning the framework itself, such as questions on "why" something was done in a particular way, suggestions and general discussions, do not use Stackoverflow - please refer to the project's issue tracker and mark your issue with the "question" tag instead. For bugs and feature requests, please also use issue tracker. If the project has been useful for you, please consider helping us improve the wiki. Thanks!