Time series models: to predict the arrival of international tourists in India

This study is to determine appropriate forecasting technique for the arrival of Tourists of Foreign countries (TFCs) in India, by adopting time series statistical tools. For the present study, data of arrival of TFCs were taken at two distinct levels; yearly arrival of FCs and Monthly arrival of TFCs. The presence of time series component like trends and seasonality is captured in present TSD (time series data). Different Exponential smoothing tools and ARIMA models were taken into the consideration for the purpose of analysing data and predicting the figures of future arrival of TFCs in India. Among the several Time series models and tools best model with appropriate tool has been explored. Various criteria were taken into the account to define best model for study like ‘Mean Absolute Percent Error’, ‘Root Mean Square Error’, ‘Akaike Information Criterion’ etc. Mehta (2015) has shown that ARIMA (0, 1, 1) is appropriate model to make prediction about TFCs, especially when observations are taken annually. This study is extension of research done by Mehta (2015). In this study, the ‘Winter's multiplicative’ model is observed to be suitable when figures for same data are taken at monthly basis.