Hybrid forecasting model of non-scheduled passenger air transportation

 In the article, an ARIMA-Fuzzy-based hybrid model is proposed for forecasting time series of non-scheduled passenger air transportation. As it is known, the ARIMA model is applied to identify linear trends and regularities within time series data as well as for forecasting. The study of scientific research literature shows that the ARIMA model has its own limitations in managing non-linearity and random changes during forecasting. Since the process of non-scheduled air transportation depends on random changes as a stochastic process, the mentioned model does not cover the whole process. For this reason, the ARIMA model does not provide effective enough results outcome strong enough to model non-linear and random changes in the data in the process of non-scheduled air transportation. In this regard, the ARIMA model was applied together with the fuzzy model. The hybrid model, based on ARIMA’s autoregression model, is applied together with the random deviation fuzzy model to further increase the accuracy of the forecast. The results obtained as a result of the application of the hybrid model show that the model in this form provides more reliable and efficient forecasts compared to independent models.