The AutoClass Project : AutoClass takes a database of cases described by a combination of real and discrete valued attributes, and automatically finds the natural classes in that data.

Autocoder Demo (text classification)

C4.5 r8 : R. Quinlan's program for top down induction of decision trees

CBA Data mining tool : CBA is a data mining tool developed at School of Computing, National University of Singapore. CBA originally stands for Classification Based on Associations. However, it turns out that it is much more powerful than simply producing an accurate classifier for prediction. It can also be used for mining various forms of association rules, and for text categorization or classification.

Con-x Connectionist Backprop Language and Simulator : Con-x (pronounced "kun ex") is a neural network scripting language and environment, designed to be used by serious backprop researchers, as well as a teaching tool for use in introductory AI courses

DELVE - Data for Evaluating Learning in Valid Experiments : Delve is a standardised environment designed to evaluate the performance of methods that learn relationships based primarily on empirical data. Delve makes it possible for users to compare their learning methods with other methods on many datasets. The Delve learning methods and evaluation procedures are well documented, such that meaningful comparisons can be made.

FastICA package for MATLAB : the FastICA package is a public-domain MATLAB program that implements the fast fixed-point algorithm for independent component analysis and projection pursuit. It features an easy-to-use graphical user interface, and a computationally powerful algorithm.

FFOIL r2 : R. Quinlan's program for inductive logic programming

FOIL r6

LIBSVM by Chih-Chung Chang and Chih-Jen Lin : LIBSVM is an integrated tool for support vector classification, (C-SVC, nu-SVC ), regression (epsilon-SVR, nu-SVR) and distribution estimation (one-class SVM ). It supports multi-class classification. The basic algorithm is a simplification of both SMO by Platt and SVMLight by Joachims. It is also a simplification of the modification2 of SMO by Keerthi et al.

MLC++ Home Page (SGI) : MLC++ is a library of C++ classes for supervised machine learning. MLC++ was initially developed at Stanford University and is now distributed by SGI.

Microsoft Belief Network Tools : an application developed by the Decision Theory Adaptive Systems Group within Microsoft Research. It allows the creation, assessment and evaluation of Bayesian belief networks.

Software by Radford Neal Available On-Line : Flexible Bayesian modeling and Markov chain sampling, Low Density Parity Check (LDPC) codes, arithmetic coding for data compression.

Neural Network Toolbox for MATLAB : this toolbox provides a complete set of functions and a graphical user interface for the design, implementation, visualization, and simulation of neural networks. It supports the most commonly used supervised and unsupervised network architectures and a comprehensive set of training and learning functions

Neural Networks at your Fingertips : simulator for Adaline, Backprop, Hopfield nets, Bidirectional Associative Memories, Boltzman Machine, Counterpropagation, Self-organizing maps, Adaptive Resonance Theory

NeuroForecaster GENETICA : full 32-bit implementation for Windows for general-purpose business and financial forecasting. Performs time-series analysis, cross-sectional classification and indicator analysis.

NEURON : NEURON is a simulation environment for developing and exercising models of neurons and networks of neurons. It is particularly well-suited to problems where cable properties of cells play an important role, possibly including extracellular potential close to the membrane), and where cell membrane properties are complex, involving many ion-specific channels, ion accumulation, and second messengers. It evolved from a long collaboration between Michael Hines and John W. Moore at the Department of Neurobiology, Duke University.

The NICO ANN Toolkit Home Page : the NICO Toolkit is an artificial neural network toolkit designed and optimized for speech technology applications. It is easy to construct neural networks with both recurrent connections and/or time-delay windows to capture temporal features. The network topology is very flexible -- any number of layers is allowed, and layers can be arbitrarily connected. Powerful tools for sparse connectivity are also included. Tools for extracting input-features from the speech signal are also part of the toolkit, as well as tools for computing target values from many common phonetic label-file formats.

The NN learning algorithm benchmarking page : proper benchmarking of (neural network and other) learning architectures is a prerequisite for orderly progress in this field. In many published papers deficiencies can be observed in the benchmarking that is performed. A workshop about NN benchmarking at NIPS*95 addressed the status quo of benchmarking, common errors and how to avoid them, currently existing benchmark collections, and, most prominently, a new benchmarking facility including a results database. This page contains pointers to written versions or slides of most of the talks given at the workshop plus some related material. The page is intended to be a repository for such information to be used as a reference by researchers in the field.

The NNCTRL Toolkit. Neural networks for control : the NNCTRL toolkit is a set of tools for design and simulation of control systems based on neural networks. The toolkit is an add-on to the NNSYSID toolbox, which is a toolbox for system identification with neural networks. Version 2 requires MATLAB 5.3 or higher. For MATLAB 4.2-MATLAB 5.2 it is possible to use the old Version 1. The toolkit contains: Control by feedback linearization. Direct inverse control. Internal model control. Optimal control. Control using instantaneous linearization (includes approximate pole placement, approximate minimum variance and approximate GPC control). Nonlinear Generalized Predictive Control. Nonlinear Feedforward Control.

PDP++ Home Page : the PDP++ software is a neural-network simulation system written in C++. It represents the next generation of the PDP software originally released with the McClelland and Rumelhart "Explorations in Parallel Distributed Processing Handbook", MIT Press, 1987. It is easy enough for novice users, but very powerful and flexible for research use.

Old PDP package

The Perceptron : a simple simulator for the Perceptron learning rule

PlaNet5.7

Pygmalion

R: the Comprehensive R Archive Network : R is `GNU S', a freely available language and environment for statistical computing and graphics which provides a wide variety of statistical and graphical techniques: linear and nonlinear modeling, statistical tests, time series analysis, classification, clustering, etc.

RuleQuest Research Data Mining Tools (C5.0, Magnum Opus)

SNNS- Stuttgart Neural Network Simulator

Spike-neuralog

SOM_PAK, LVQ_PAK

SOM Toolbox for Matlab

SUBDUE Knowledge Discovery in Structural Databases

StatLog: Evaluation - Characterization of Classification Algorithms : this work was supported by Esprit Project 5170 StatLog (1991-94). This project was concerned with comparative studies of different machine learning, neural and statistical classification algorithms. About 20 different algorithms were evaluated on more than 20 different datasets. The tests carried out under this project produced many interesting results. Site contains datasets and algorithms

SVM-Light Support Vector Machine : SVM^light is an implementation of Support Vector Machines (SVMs) in C written by T. Joachims

TiMBL: Tilburg Memory Based Learner : TiMBL is a program implementing several Memory-Based Learning techniques. TiMBL stores a representation of the training set explicitly in memory, and classifies new cases by extrapolation from the most similar stored cases. Several metrics and algorithms are implemented in TiMBL; among others: Information Gain weighting for dealing with features of differing importance (IB1-IG), and the Modified Value Difference metric for making graded guesses of the match between two different symbolic values. TiMBL is optimized for fast classification by using several indexing techniques and heuristic approximations (such as IGTREE and TRIBL).

Tlearn software page

UC Irvine KDD Archive

UC Irvine Machine Learning Datasets Repository

UC Irvine Machine Learning Programs

WEKA Machine Learning Project : several standard ML techniques into a software "workbench" called WEKA, for Waikato Environment for Knowledge Analysis

top

29th April 2003. Machine Learning and Neural Networks Group. For questions and comments: aceroni@dsi.unifi.it.