ARIMA based time-series forecasting technique for daily prediction of stock market indices

The study presents an autoregressive integrated moving average (ARIMA) based time-series forecasting technique for prediction of stock market indices. The model has been considered for daily prediction of Tata Motors stock considering data of past one year from the Bombay Stock Exchange (BSE), India portal. The model has been trained in python-spyder platform and analysed for predicting next 10 days of stock values in terms of open price, high price, low price and closing price. The performance of the ARIMA model has been assessed in terms of different error indicators namely mean absolute error (MAE), root mean square error (RMSE) and mean absolute percentage error (MAPE). The results confirmed that the model can precisely predict the output with minimal error values. The results also show that the prediction error increases with duration of prediction.