Towards Forecasting Business Prepaid Mobile Telecommunication Using Connectionist Model

Prepaid mobile service has become a necessity to the society and contributed success to many business. Realizing its importance, the mobile networks are moving towards higher data rates and packet oriented data transmission and mobile having more multimedia features. This fact has open end opportunities for new and more attractive mobile technologies. However, forecasting the business trend in this domain is a difficult task as it involves time dependency data. Hence, this study proposed a connectionist model as an alternative for forecasting mobile business trend. In this study, the teletraffic data was gathered from Celcom Service Control Point (SCP). Neurall network was trained with SCP data to forecast Celcom mobile business trend. This result will help Celcom in their business planning. This study has proven the capability and reliability of the connectionist model in performing the forecasting business prepaid mobile telecommunication. The performance of the back propagation model with the accuracy above 97 percent is satisfactory. The model is able to capture data from the call event records towards forecasting the trends of business prepaid mobile Telco, thus making it short and simple to use.