Predictive Modeling on Telekom Malaysia Direct Exchange Line Growth

Data mining (DM) is the extraction of hidden predictive information from large databases that has becoming a powerful new technology with great potential to help companies focus on the most important information in their data warehouses. A predictive model makes a prediction about values of data using known results found from historical data where the best possible outcome based on the previous data is derived. Telekom Malaysia (TM) is Malaysia's premier communications provider that provides the digital backbone and communication facilities. Direct Exchange Line (DEL) is one of its core telephony services that handle massive volume and variety of data in its daily operations. Therefore, it is hard to reveal knowledge structures that can guide decisions in conditions
of limited certainty. The main objective of this study is to identify the most appropriate DM technique between logistic regression, decision tree and neural networks for
predicting DEL growth based on five physical attributes namely exchanges, subscribers, new installation, cutting, and availability of cable or ports (ECP) that constitute of 672 instances leading to a target (either increase or decrease). The result of this study is important in assisting the prediction of DEL growth in TM specifically in Penang, thus leading on better understanding on the future of the market based on the current and previous situation.