Forecasting Age Adjusted Rates of Lung Cancer in Mumbai by Fitting ARIMA Models

AbstractThere are several methods to forecast the values of any time series. The best model is one which has minimum error. If the given time series is stationary, we can use the autoregressive integrated moving average (ARIMA) model. If it is not stationary, it has to be made stationary by differencing the time series. ARIMA model was identified using the Bayesian Information Criteria. The Auto Covariance Function (ACF) and Partial Auto Covariance Function (PACF) were studied along with the stationarity of the time series. The future values were forecasted and the past values were predicted using recursive method.KeywordsStationaryTime seriesAge adjusted ratesAutocorrelationAurtoregressive modelsAkaike information criterionBayesian information criterion