On Interference of Signals and Generalization in Feedforward Neural Networks

TL;DR

This paper investigates how signal interference within feedforward neural networks impacts their ability to generalize, highlighting that interference can cause unpredictable generalization, and explores adaptive activation functions to mitigate this issue.

Contribution

It introduces the concept of signal interference affecting generalization in neural networks and evaluates adaptive activation functions as a solution.

Findings

Interference can lead to highly random generalization.

Adaptive activation functions can reduce unpredictable generalization.

Experimental tests confirm the impact of interference on neural network performance.

Abstract

This paper studies how the generalization ability of neurons can be affected by mutual processing of different signals. This study is done on the basis of a feedforward artificial neural network. The mutual processing of signals can possibly be a good model of patterns in a set generalized by a neural network and in effect may improve generalization. In this paper it is discussed that the interference may also cause a highly random generalization. Adaptive activation functions are discussed as a way of reducing that type of generalization. A test of a feedforward neural network is performed that shows the discussed random generalization.