Stock Forecasting using Fuzzy Neural Networks, Technical Indicators, and Foreign Exchange Rates

Abstract

Using features derived from technical indicators of stocks and foreign exchange rates, this paper presents a forecasting plan for the KOSPI's rise and fall after one day based on the neural network with weighted fuzzy membership functions (NEWFM). To extract the features to be used in NEWFM, select the three technical indicators in the first step: Relative Strength Index (RSI), Commodity Channel Index (CCI), and Current Price Change (CPC). The second phase extracts 13 features derived from the KOSPI data from the three technology indicators selected in the first stage. It also extracts one feature derived from the technical indicators of the KRW/USD exchange rates from the KRW/USD exchange rates. In this way, 14 features will be used as inputs to NEWFM to forecast the KOSPI's rise and fall in a day. NEWFM uses 13 and 14 features to show forecasting performance of 58.86% and 59.38%, respectively. These experiments show that the KRW/USD exchange rates affects the rise and fall of KOSPI stock prices. © 2020, World Academy of Research in Science and Engineering. All rights reserved.