Time Series Forecasting

Forecasting is important in economics, commerce, and in various disciplines of social science and pure science. Forecasting is a method for computing future values by analysing the behaviour of present and past values of a time series. Forecasting model may be univariate or multivariate. In the univariate model, forecasts depend on present and past values of the single time series being forecasted. In a multivariate model, forecasts of a time series variable depend on values of one or more explanatory variables. This chapter aims to provide an overview of forecasting based on time series analysis. Forecasting on time series is essentially a form of extrapolation, which involves estimating a model with a sample dataset and using the estimated model outside the range of data by using which the model has been estimated.