Prediction and Forecasting of Automotive Sale...

Prediction and Forecasting of Automotive Sales in India Using ARIMA models

Author Name : Dr. Sameerabanu P

ABSTRACT

A time series data  of yearly annual Automotive  sales  is analysed. These Statistics are market volumes, which include passenger cars as well as light commercial vehicles (LCVs).This Time series Data is composed of 16 year observations. The importance of this study came from the application of Box and Jenkins (ARIMA - Auto Regressive Integrated Moving Average) approach on the analysis of this yearly series data using the SPSS software program. More specifically, the time series is examined and the appropriate ARIMA model is identified for these data and the models parameters are estimated. After that, diagnostic check of the residuals is performed for autocorrelation, variance and normality assumptions. Finally, the appropriate ARIMA model is used for forecasting, so that decision-makers will rely on it to make future plans.

Keywords: ARIMA, Autocorrelation, Partial Autocorrelation, MAPE (Mean Absolutely Percentage Error),  RMSE (Root Mean Square Error ) and Normalised BIC ( Normalized Bayesian Information criterion).