About: The SHOGUN machine learning toolbox's focus is on large scale learning methods with focus on Support Vector Machines (SVM), providing interfaces to python, octave, matlab, r and the command line. Changes:This release features the work of our 8 GSoC 2014 students [student; mentors]:
It also contains several cleanups and bugfixes: Features
Bugfixes
Cleanup and API Changes

About: Distributed optimization: Support Vector Machines and LASSO regression on distributed data Changes:Initial Upload

About: The fertilized forests project has the aim to provide an easy to use, easy to extend, yet fast library for decision forests. It summarizes the research in this field and provides a solid platform to extend it. Offering consistent interfaces to C++, Python and Matlab and being available for all major compilers gives the user high flexibility for using the library. Changes:Initial Announcement on mloss.org.

About: Hubnessaware Machine Learning for Highdimensional Data Changes:

About: A template based C++ reinforcement learning library Changes:Initial Announcement on mloss.org.

About: C++ generic programming tools for machine learning Changes:Initial Announcement on mloss.org.

About: Java package for calculating Entropy for Machine Learning Applications. It has implemented several methods of handling missing values. So it can be used as a lab for examining missing values. Changes:Discretizing numerical values is added to calculate mode of values and fractional replacement of missing ones. class diagram is on the web http://profs.basu.ac.ir/bathaeian/free_space/jemla.rar

About: 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 productiongrade 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 online documentation helps fill in the details. Changes:Adding a large number of new distributions, such as AndersonDaring, ShapiroWilk, Inverse ChiSquare, Lévy, Folded Normal, Shifted LogLogistic, Kumaraswamy, Trapezoidal, Uquadratic and BetaPrime distributions, BirnbaumSaunders, Generalized Normal, Gumbel, Power Lognormal, Power Normal, Triangular, Tukey Lambda, Logistic, Hyperbolic Secant, Degenerate and General Continuous distributions. Other additions include new statistical hypothesis tests such as AndersonDaring and ShapiroWilk; as well as support for all of LIBLINEAR's support vector machine algorithms; and format reading support for MATLAB/Octave matrices, LibSVM models, sparse LibSVM data files, and many others. For a complete list of changes, please see the full release notes at the release details page at: https://github.com/accordnet/framework/releases

About: a parallel LDA learning toolbox in MultiCore Systems for big topic modeling. Changes:Initial Announcement on mloss.org.

About: Gaussian processes with general nonlinear likelihoods using the unscented transform or Taylor series linearisation. Changes:Initial Announcement on mloss.org.

About: LuaMapReduce framework implemented in Lua using luamongo driver and MongoDB as storage. It follows Iterative MapReduce for training of Machine Learning statistical models. Changes:

About: An extensible C++ library of Hierarchical Bayesian clustering algorithms, such as Bayesian Gaussian mixture models, variational Dirichlet processes, Gaussian latent Dirichlet allocation and more. Changes:Initial Announcement on mloss.org.

About: LogRegCrowds is a collection of Julia implementations of various approaches for learning a logistic regression model multiple annotators and crowds, namely the works of Raykar et al. (2010), Rodrigues et al. (2013) and Dawid and Skene (1979). Changes:Initial Announcement on mloss.org.

About: This library implements the OptimumPath Forest classifier for unsupervised and supervised learning. Changes:Initial Announcement on mloss.org.

About: pySPACE is the abbreviation for "Signal Processing and Classification Environment in Python using YAML and supporting parallelization". It is a modular software for processing of large data streams that has been specifically designed to enable distributed execution and empirical evaluation of signal processing chains. Various signal processing algorithms (so called nodes) are available within the software, from finite impulse response filters over datadependent spatial filters (e.g. CSP, xDAWN) to established classifiers (e.g. SVM, LDA). pySPACE incorporates the concept of node and node chains of the MDP framework. Due to its modular architecture, the software can easily be extended with new processing nodes and more general operations. Large scale empirical investigations can be configured using simple text configuration files in the YAML format, executed on different (distributed) computing modalities, and evaluated using an interactive graphical user interface. Changes:improved testing, improved documentation, windows compatibility, more algorithms

About: BayesOpt is an efficient, C++ implementation of the Bayesian optimization methodology for nonlinearoptimization, experimental design and stochastic bandits. In the literature it is also called Sequential Kriging Optimization (SKO) or Efficient Global Optimization (EGO). There are also interfaces for C, Matlab/Octave and Python. Changes:Fixed bugs and doc typos

About: Multicore nonparametric and bursty topic models (HDPLDA, DCMLDA, and other variants of LDA) implemented in C using efficient Gibbs sampling, with hyperparameter sampling and other flexible controls. Changes:Corrections to diagnostics and topic report. Correction to estimating alpha. Now estimating beta sometimes (when estimating phi).

About: A Content Anomaly Detector based on nGrams Changes:Lots and lots of cool new features and bugfixes ;)

About: The implementation of adaptive probabilistic mappings. Changes:Initial Announcement on mloss.org.

About: Boosting algorithms for classification and regression, with many variations. Features include: Scalable and robust; Easily customizable loss functions; Oneshot training for an entire regularization path; Continuous checkpointing; much more Changes:
