:: data-machine ::  
:: d a t a : m a c h i n e :: 
tools for computational intelligence
Artificial Neural Networks

   

Home

Downloads

Topics:

Introduction

Neural networks

Fuzzy systems

Chaotic systems

Global
and evolutionary
optimization

To receive
information about
the new updates
please send an email:


 

You are invited to try NeuralMachine - Neural network tool

Introduction

Originally studied in the framework of AI, artificial neural networks (ANNs) has become now one of the primary technologies in machine learning, and a mainstream technology for data-driven modelling. Various types of networks are used in clustering, classification and prediction. ANNs loosely imitate functioning of neurons in a human's brain, and it appeared it is possible to combine these neurons in such a way, that the network would reproduce any multi-variable multi-valued function, given enough points and values of this function. By analogy with the brain, the operation of the trained (learned) network is often called recall.

ANNs exhibit three features, namely, distributed processing, adaptation and nonlinearity, and it have been mathematically proven that adding up simple functions, as a ANN does, allows for universal approximation of functions (Kolmogorov 1957). This means that neural networks can approximate any function that best characterizes a time series. It is this property that has stimulated civil engineers to adapt, investigate, and improve the performance of neural networks associated with their applications.

Read more on ANNs in the NeuralMachine Manual (opens in another tab or window).

Other useful resources:

Neural Network FAQ, part 1 of 7, by Warren S. Sarle
Bibliographies on Neural Networks
Links to Neural Nets Research