Wolfram Mathematica Neural Networks

Category Intelligent Software>Neural Network Systems/Tools

Abstract Wolfram Mathematica Neural Networks (WMNN) gives professionals and students the tools to train, visualize, and validate neural network (NN) models.

It supports a comprehensive set of neural network structures-- including radial basis function, feed-forward, dynamic, Hopfield, Perceptron, vector quantization, unsupervised, and Kohonen networks.

It implements state-of-the-art training algorithms like Levenberg- Marquardt, Gauss-Newton, and steepest descent.

WMNN also includes special functions to address typical problems in data analysis, such as function approximation, classification and detection, clustering, nonlinear time series, and nonlinear system identification problems.

WMNN is equally suited for advanced and inexperienced users.

The built-in palettes facilitate the input of any parameter for the analysis, evaluation, and training of your data.

The online documentation contains a number of detailed examples that demonstrate different neural network models.

You can solve many problems simply by applying the example commands to your own data.

WMNN also provides numerous options to modify the training algorithms. The default values have been set to give good results for a large variety of problems, allowing you to get started quickly using only a few commands.

As you gain experience, you will be able to customize the algorithms to improve the performance, speed, and accuracy of your neural network (NN) models.

With WMNN and Mathematica (see Note 1), you will have access to a robust modeling environment that lets you test and explore neural network (NN) models faster and easier than ever before.

The package comes with extensive printed and electronic documentation.

Wolfram Mathematica Neural Networks features/capabilities include:

Easy to Use and Learn --

1) Small number of functions constructed so that only the minimum amount of information has to be specified by the user.

2) Well-organized palettes with command templates, options, and links to online documentation.

3) Intelligent initialization algorithms to begin the training with good performance and speed.

4) Extensive documentation including an introduction to neural network (NN) theory as well as highly illustrative application examples.

Support for Proven Neural Network (NN) Paradigms --

1) Support for most of the commonly used neural network (NN) structures including radial basis function, feed-forward, dynamic, Hopfield, Perceptron, vector quantization, unsupervised, and Kohonen networks.

2) Support for advanced training algorithms including Levenberg- Marquardt, Gauss-Newton, and steepest descent as well as for traditional algorithms including Backpropagation with and without momentum.

3) Support for typical neural network (NN) applications including function approximation, classification, dynamic systems modeling, time series, auto-associative memory, clustering, and self-organizing maps.

Advanced Modeling Environment --

1) Visualization tools for viewing network models, the training process, and network performance.

2) Special 'network object' to identify the type of network and list its parameters and properties.

3) Special training record to keep intermediate information from the learning process.

4) Functions equipped with a large number of 'advanced options' to modify and control the training algorithms.

5) Support for neural networks (NN) with any number of hidden layers and any number of neurons (hidden neurons) in each layer.

6) Access to all of the capabilities of Mathematica to prototype new algorithms or to perform further manipulations on neural network (NN) structures.

Fast and Reliable --

1) Optimization of expressions before numerical evaluation to minimize the number of operations and to reduce computational load.

2) Compile command to send compiled code directly to Mathematica to increase computational speed.

3) Special performance-evaluation functions included to validate and illustrate the quality of a mapping.

Note 1: Mathematica is a computer program used mainly in scientific, engineering and mathematical fields. It was originally conceived by Stephen Wolfram and developed by a team of mathematicians and programmers that he assembled and led. It is sold by Wolfram Research and its distributors.

Mathematica provides cross-platform support for tasks such as symbolic or numerical calculations, arbitrary precision arithmetic, data processing, and plotting. Mathematica offers a programming language which supports functional and procedural programming styles.

System Requirements

Wolfram Mathematica Neural Networks 1.0.2 requires Mathematica 5.0.1 - 5.2 and is available for all Mathematica platforms.

Mathematica platforms


Manufacturer Web Site Wolfram Mathematica Neural Networks

Price Contact manufacturer.

G6G Abstract Number 20394

G6G Manufacturer Number 102980