PENYUSUNAN COMPOSITE LEADING INDICATOR SIKLUS BISNIS DI INDONESIA MENGGUNAKAN ARTIFICIAL NEURAL NETWORK

Abstract

After  the 1997  financial  crisis,  like many  developing  Asian countries,  Indonesiahave taken major initiatives to improve their national statistical systems as part oftheir efforts  in  strengthening national  economic monitoring  and  surveilance  andcrisis prevention measure.  Composite leading indicators are becoming more widelyrecognized in predicting business cycles in Indonesia.This article attemps to construct composite leading indicators of business cycles inIndonesia using monthly economic and financial data during the sample period of1970–2001.    This  article  also  explore  the  possibility  of  constructing  compositeleading indicators of business cycles by using artificial neural network method.  Theresult show  that  the composite  leading  indicators constructed by artificial neuralnetwork method  is  able  to  predict  all of  the  turning points  of  business cycles  inIndonesia.  The performance of this method is comparable to the other predictionmethod such as regression.

Keywords: Aggregate economic activity, business cycles, artificial neural network,predicting, composite leading  indicators, turning points