ANALISIS CONDITIONAL RESTRICTED BOLTZMAN MACHINE UNTUK MEMPREDIKSI HARGA SAHAM BANK SYARIAH INDONESIA

Abstract

This study aims to predict the movement of Bank Syariah Indonesia shares (BRIS.JK) prices using the Conditional Restricted Boltzmann Machine (CRBM) method. Prediction is needed in conducting share transactions, because the increase or decrease in share price movements is very difficult to predict. The CRBM method is a machine learning algorithm used to model the probability distribution of data associated with variables and inputs. CRBM is a type of Restricted Boltzman Machine (RBM) that consists of two layers, namely the input layer and the hidden layer. CRBM is a type of Boltzmann machine model equipped with a conditioned unit that is used to perform analysis and learning on data that has conditional properties. In this research, the first step is to divide several research scenarios. Then conduct CRBM tests to get prediction results. The data used is daily close data. Based on the research that has been done, it is obtained that the best prediction accuracy is in July - August with MAPE below 5%.