Quality Solution of Logic Programming in Hopfield Neural Network

Abstract

The dynamical behaviours of Artificial neural network (ANN) system are strongly dependent by its network structure. In that sense, the output of ANN has long suffered from a lack of interpretability and variation. This has severely limited the practical usability of ANN in doing logic programming. The work presents an integrated representation of 2 Satisfiability (2SAT) in different Hopfield Neural Network (HNN) models. Neuron states of HNN always converge to minimum energy but the solution produced always confined in limited number of solution space. The main purpose of this study is to explore the quality of the solution obtained from HNN. It has been shown that HNN only retrieves limited neuron states with the lowest minimum energy. This finding will lead to a better understand of logic programming in HNN.