A free from local minima algorithm for training regressive MLP neural networks

TL;DR

This paper introduces a novel training algorithm for regressive MLP neural networks that avoids local minima by leveraging the distribution of the training set, demonstrated on a standard benchmark.

Contribution

The paper presents a new training method that circumvents local minima issues in MLP networks by utilizing the training set's distribution properties.

Findings

Algorithm successfully avoids local minima

Effective on a well-known benchmark

Outperforms traditional backpropagation methods

Abstract

In this article an innovative method for training regressive MLP networks is presented, which is not subject to local minima. The Error-Back-Propagation algorithm, proposed by William-Hinton-Rummelhart, has had the merit of favouring the development of machine learning techniques, which has permeated every branch of research and technology since the mid-1980s. This extraordinary success is largely due to the black-box approach, but this same factor was also seen as a limitation, as soon more challenging problems were approached. One of the most critical aspects of the training algorithms was that of local minima of the loss function, typically the mean squared error of the output on the training set. In fact, as the most popular training algorithms are driven by the derivatives of the loss function, there is no possibility to evaluate if a reached minimum is local or global. The algorithm presented in this paper avoids the problem of local minima, as the training is based on the properties of the distribution of the training set, or better on its image internal to the neural network. The performance of the algorithm is shown for a well-known benchmark.