Document worth reading: “A Review of Modularization Techniques in Artificial Neural Networks”

Artificial neural networks (ANNs) have achieved vital success in tackling classical and trendy machine finding out points. As finding out points develop in scale and complexity, and improve into multi-disciplinary territory, a further modular technique for scaling ANNs might be wished. Modular neural networks (MNNs) are neural networks that embody the concepts and guidelines of modularity. MNNs undertake an enormous amount of completely completely different methods for reaching modularization. Previous surveys of modularization methods are comparatively scarce in their systematic analysis of MNNs, focusing completely on empirical comparisons and lacking an intensive taxonomical framework. In this overview, we intention to determine a secure taxonomy that captures the vital properties and relationships of the completely completely different variants of MNNs. Based on an investigation of the completely completely different ranges at which modularization methods act, we attempt to provide a typical and systematic framework for theorists studying MNNs, moreover attempting alongside one of the simplest ways to emphasize the strengths and weaknesses of completely completely different modularization approaches in order to concentrate on good practices for neural group practitioners. A Review of Modularization Techniques in Artificial Neural Networks