Forecasting of Immigrants in Canada using Forecasting models

Abstract

In Canada, the number of international students, temporary workers and refugees from every part of the world grows each year. Therefore, forecasting immigration is important for the economy of Canada and Labor Market. In this regard, four forecasting approaches have been applied to the annual data of immigrants for the period 2000-2019. The accuracy of Moving average (MA), Autoregressive (AR), Autoregressive moving average (ARMA), Autoregressive integrated moving average (ARIMA) models were checked via comparing Akaike’s information criteria(AIC), Bayesian information criteria (BIC), Mean error (ME), Root mean square error(RMSE), Mean absolute error (MAE), Mean percentage error (MPE), Mean absolute percentage error (MAPE) and Mean absolute scaled error (MASE) and graphical approaches such as ACF plots of residuals. Experimental results showed that ARIMA (1,2,4) is the best-fitted model for forecasting immigrants in Canada. Selected forecasting approaches are applied to predict immigrants for five years from 2020-2024.