Growing Neural Gas approach for obtaining homogeneous maps by restricting the insertion of new nodes

Abstract

The Growing Neural Gas model is used widely in artificial neural networks.However, its application is limited in some contexts by the proliferation of nodes in dense areas of the input space. In this study, we introduce some modifications to address this problem by imposing three restrictions on the insertion of new nodes. Each restriction aims to maintain the homogeneous values of selected criteria. One criterion is related to the square error of classification and an alternative approach is proposed for avoiding additional computational costs.Three parameters are added that allow the regulation of the restriction criteria.The resulting algorithm allows models to be obtained that suit specific needs by specifying meaningful parameters.

A new neural gas network approach for obtaining the singularity-free workspace of 3-DOF planar parallel manipulators

“…Since then, it has been mostly applied in the context of vector quantization or data compression, such as image processing, speech recognition or pattern recognition. 30,31 A batch variant of NGN which shows much faster convergence and which can be interpreted as an optimization of the cost function by the Newton method was proposed by Cottrell et al 32 Quintana-Pacheco et al 33 proposed a modified version of the NGN algorithm where the main goal consists of avoiding prototype proliferation in dense areas of the input space. Garcı´A-Rodrı´Guez et al 34 addressed the ability of self-organizing neural network models to manage real-time applications by introducing the so-called fAGNG (fast Autonomous Growing Neural Gas), a modified learning algorithm for the incremental model NGN network.…”

“…32 Quintana-Pacheco et al. 33 proposed a modified version of the NGN algorithm where the main goal consists of avoiding prototype proliferation in dense areas of the input space. GarcíA-RodríGuez et al.…”