A GENETIC AND NEURAL NETWORK APPROACH TO SIGNIGICANT INDICATORS SELECTION

Abstract

The determination of significant indicators to forecast the value of a stock or mutual fund has been a tough and confusing problem. In this paper, we propose a fund net asset value (NAV) forecasting model, which combines a neural network and a generic algorithm, providing useful investment reference to investors. First, we use our proposed model to find out the significant indicators for forecasting the NAVs of mutual funds. Next, we use those significant indicators to forecast the NAVs of mutual funds. According to the experiments on Taiwan mutual funds, we found that there did exist common significant indicators for forecasting the NAVs of mutual funds.