Forecasting Stock Market Indices Using Hybrid Network

In this paper, a hybrid network consisting of a trigonometric Functional Link Artificial Neural Network (FLANN) and Fuzzy Logic System named as Functional Link Neural Fuzzy (FLNF) Model is used to predict the stock market indices. The proposed model uses a functional link neural network to the consequent part of the fuzzy rules. The consequent part of FLNF model is a non-linear combination of input variables. Two stock market indices (data sets) i.e., Bombay Stock Exchange and Standard's and Poor's (S&P500) are collected for experimentation. Samples for 4000 trading days from 1(st) March 1993 to 23(rd) July 2009 are collected from the former and 3228 trading days from 1(st) March 1993 to 09(th) June 2006 for the later. This model is used to forecast stock market indices one day, one week and one month in advance. A comparative analysis between the proposed hybrid model and that of FLANN has also been given. The MAPE and RMSE are used to find out the performance of both the models and it shows the superiority of the hybrid model.