Computers

Short article on self-modifying software touching on co-training, partial-programming, and genetic programming (GP) methods of problem solving. [FREE]

Programs to cluster similar contexts together using unsupervised knowledge-lean methods for word sense discrimination, email categorization, and name discrimination. Written in Perl. [GNU]

A resource for open source software at NASA

Various sorting algorithms including insertion, quick, merge, heap, Dutch National Flag, and radix with on-line demos. [Free]

A program which discovers interesting and repetitive subgraphs in labeled graph representations using the minimum description length principle. Includes applications to molecular biology. [Free]

A Highly Scalable, Distributed Open Source Storage Database providing index free adjacency. [AL]

This package forms a complete gradient descent machine learning library. Modules support vector machines in classification and regression, ensemble models such as bagging or adaboost, non-parametric models such as K-nearest neighbors, Parzen regression, and Parzen density estimation. Includes speech recognition tools. Written in C++ [BSD]

Software which allows one to navigate (fly) through the data tree, zoom in on interesting nodes, click on bars to get counts, and mark interesting places in the tree. Includes datasets for automobiles, voting, produce, and medical research. Uses LEDA, ([AFL] licensed only). [GPL]

This tool implements Hidden Markov Models and application to part-of-speech tagging. Also available; a multivariate hypothesis testing software for gaussian data, and a groundtruth/metadata editing and visualizing toolkit for OCR. [GPL]

A software package which allows variance-modeled posterior inference to be performed automatically on a Bayesian network. [GPL]

A collection of tools that implement decision trees and tables, rule learners, Naive Bayes, support vector machines, voted perceptrons, multi-layer perceptron. Meta schemes include bagging, stacking, and boosting. Written in Java. [GPL]

A set of tools for Windows 2000/NT/XP that allow you to build statistical models from data. [Free]

Small portable online hand recognition system based on Support Vector Machines. Provides a relatively small machine model running at 50-100 char per second recognition speed. [BSD]

This library allows probabilistic sequence models to be constructed through use of Hidden Markov models (HMMs) and Hierarchical Markov models HMMs (HHMMs) in Ocaml programming language. [GPL]

Programmatically isolate similarities between scattered classes of genes. Expression driven. Utilizes a voting method along with a k-Nearest-Neighbors classification. Very rich graphical interface. Samples of an unknown class are possible given enough data. Fully functional demo. [Commercial]

A database of cases described by a combination of real and discrete valued attributes, and automatically finds the natural classes in that data. It can be seen as a Naive Bayes classifier where the class node is hidden. [Free]

Supports several inference algorithms and learning algorithms. Allows simulation of static and dynamic networks, including HMMs, IOHMMs, and Kalman filters. [GPL]

A library of C++ classes for Bayesian Filtering of discrete systems. [MIT]

A windows-based program that classifies text based on trained material. Designed for automated essay scoring, BETSY can be applied to any text classification task. [GPL]

A library of C code useful for writing statistical text analysis, language modeling, and information retrieval programs. The current distribution includes the library, as well as front-ends for document classification (rainbow), document retrieval (arrow) and document clustering (crossbow). [LGPL]

The home page of R. Quinlan with FTP links to FOIL (inductive logic programming) and C4.5 (learning decision trees). [LGPL]

A general approach to the problem of inducing natural language parsers. It uses an annotated corpus, and produces a parser by using ILP for inducing the rules that control the actions of a shift-reduce parser. [Free]

The Carnegie Mellon University School of Computer Science select list of a few very good machine learning systems. A list of 4 FTP repository links is also listed for exploring. [GPL]

A concise explanation of backward propagation neural networks for handwriting analysis. Beginning functions are given for its 'Brain Net' library in VB.Net. [LGPL]

Computational model of human language acquisition written in Java; currently acquires a protolanguage of nouns and verbs language based on visual perception. [BSD]

Generates Gaussian mixture models for large datasets using efficient KD-clustering algorithms. [Free]

Fuzzy machine learning framework is a library of ADA packages and a GUI front-end based on graph-schemes, intuitionistic fuzzy sets and the possibility theory. Sources can be used on any platform where an Ada 2005 compiler is available. [GPL]

A toolset of genetic algorithm objects for C++ to perform optimization. Uses any representation and genetic operators. The documentation contains implementation and examples. Nice screenshots. PVM for distributed, parallel implementations. Includes graphic examples that use the Athena, Motif widget sets, or MFC. [BSD]

Hidden Markov Models software library from the Center of Applied Informatics, Cologne. Includes algorithms such as Viterbi, Baum-Welch, and Forward-Backward. [GPL]

A large collection of algorithm projects with complete sourcecode. [GPL]

Provides GMDH-based machine learning technology for classification, continuous value prediction and time series forecasting. The software uses multi-core processors and HPC Linux clusters. [Commercial]

A fast graph clustering software that computes normalized cut and ratio association for a given undirected graph without any eigenvector computation.

A large collection of Intelligent Agent projects complete with sourcecode. [GPL]

An algorithm that incrementally constructs decision trees from labeled examples. [AFL]

Many algorithms, toolboxes and full languages are available for scientific computing. Abstracts, PDF, and source code is available. [GPL]

JProGraM is a machine learning library which supports learning and inference algorithms for Bayesian networks, Markov random fields, hybrid random fields, probabilistic decision trees, dependency networks, and Parzen windows. [GPL]

The aim of this project is to develop a Computational Environment for integrating the design and use of knowledge extraction models from data using evolutionary algorithms. Genetic learning may also be applied to the model. [GPL]

A formal system for constructing artificial intelligent software agents from the book, 'Design of Logic-based Intelligent Systems' covering such topics as production rules, neural nets, support vector machines, fuzzy logic, and Bayesian networks. License is given through purchase of the book alone. [COMM]

A library of classes for optimizing (training) the generic models. Written in C++. Permission for use is required outside of classroom. [GPL]

A general purpose library for Bayesian filtering written in C++. [LGPL]

A machine learning library. New implementations of various machine learning algorithms. [GPL]

A suite of Java libraries for the linguistic analysis of human language which can link entity mentions to database entries, uncover relations, cluster documents, and discover significant trends. [GPL]

A production rule based system implemented in a Common Lisp Object System (CLOS) for the development of Lisp based intelligent software agents. [LGPL]

A collaborative machine learning weblog by John Langford.

Various classes, packages, macros and other software systems related to machine learning. [GPL]

An integrated collection of Java code useful for statistical natural language processing, document classification, clustering, information extraction, and other machine learning applications to text. [GPL]

A Markov Logic Interpreter that focuses on efficient MAP inference and Online Learning featuring MAP inference using Cutting Planes combined with Max-Walk-Sat programming, parametrized weights, a shell interpreter, and cardinality constraints. [GPL]

A community effort listing of reproducible research via open source software, open access to data and results, and open standards for interchange. [FREE]

A high performance Python package for predictive modeling. Fast N-dimentional array manipulation is performed via numpy using C code. New features include: OLS, Ridge Regression, Kernel Redge Regression, LASSO, LARS, Gradient descent for Regression, and K-Means. [GPL]

A generic framework for the evolutionary search algorithm mPOEMS. It was designed to solve optimisation problems, with an unrestricted number of objectives. This site provides all sources and some exemplary implementations, e.g. of the n-hard knapsack problem. [AFL]