The ANFIS System for Nonlinear Combined Forecasts in the Telecommunications Industry
The Adaptive Network-Based Fuzzy Inference System (ANFIS) has been proven to be efficient for forecasting. To address this concern, this research develops a nonlinear combined forecasting system by ANFIS for predicting the demand of telecommunication technology. The authors investigate the weights assigned to the combined forecast using two linear methods (the Least squares analysis and the Logistic model), as well as one nonlinear methods (the Bass model). An empirical data set from 3G technology development in Taiwan is used to demonstrate the application of the proposed methodology. These results show that the ANFIS method outperforms other individual methods. Also, this proposed work also provides the user with a user interface in which user can fill the query and find the desired forecasting results.