Penggunaan Sifat Pengingat Asosiatif Pada Jaringan Syaraf Tiruan Hopfield Diskret Untuk Pemulihan Data

Abstract

This research concern with application of discrete Hopfield neural networks for recovering data. Using associative memories properties of discrete Hopfield neural networks we can store a set of data patterns as a memories. The basic concepts of using discrete Hopfield neural networks as associative memories is to interpret the system’s neurons evolution as a movement of input pattern toward the one stored pattern most resembling the input pattern. The result shows that the application could recover false data to its origin.