Predicting is very hard for anything particularly if a relationship between inputs and results are non-linear in characteristics and artificial neural networks are one of them. Forecasting stock data with a customary time series evaluation has shown to be intricate. Because no primary postulation about a suitable mathematical model must be made prior to prediction furthermore, a neural network gets the knack to draw out useful information from significant collections of data, which frequently is necessary for a gratifying description of the financial time series.