Softcomputing Approach for Stock Price Trend Forecasting from Multivariate Time Series

Forecasting of stock market price is an important as well as very difficult problem. Stock market price depends on various factors and their complex relationships. In this sional time stamped data with complex relation ships. Neu ral networks are good for function approximation and can be used for predicting future values of a time series from its past values. In our case we need to handle multidimen work a hybrid soft computing approach based on the theory sional time series data and use of neural network for high of rough set and artificial neural network has been developed to predict the trend of stock market values of a particular company from its past data and also past values of various other influencing financial time series data. The simulation results from two sets of data shows on average around 70% accuracy of trend prediction.