Abstract: Artificial neural networks (ANNs) are a flexible computing frameworks and universal approximates that can be applied to a wide range of time series forecasting problems with a high degree of accuracy for the convenience of predicting the futuristic value in share market and give a better future scope for investment. But yet the artificial neural network is not to the satisfactory as it includes both theoretical and empirical findings have concluded that the combination of different models can be an effective way of improving upon the predictive performance, if the models in ensemble are quite different. In this paper, a novel hybrid model of artificial neural network is proposed using a auto-regressive integrated moving average (ARIMA) model to produce the more accurate forecasting model than artificial neural network. On this context, we collected data on monthly closing stock indices of sensex, on these we have tried to develop an appropriate model that would help us to forecast the future unknown values of Indian stock market indices, i.e, ARIMA. Therefore, it can be used as an appropriate alternative model for forecasting task, especially when higher forecasting accuracy is needed.

Keywords: Sensex, Time Series, ARIMA model, validation.