An efficient framework of predictive data mining under a case study of gas energy production in pakistan

TLDR: This paper proposed a framework for developing a predictive data mining system, which will efficiently work in scenario of forecasting with an efficient flow of work and showed twelve steps and their sequence of working in the form of algorithm.

Abstract: A lot of research has been carried out to study the energy crisis in Pakistan by using the predictive data mining techniques. Many researchers have tried to analyse the situation by using different frameworks but unfortunately the authors of this paper did not find any complete, cost effective and efficient framework in the literature. We therefore proposed a framework for developing a predictive data mining system, which will efficiently work in scenario of forecasting with an efficient flow of work. In this conceptual framework authors have showed twelve steps and their sequence of working in the form of algorithm. We applied the proposed framework on the case study for prediction of natural gas energy production in Pakistan to give a solution of the current energy crises. In this case study efforts were made to collect the historical data by covering different geographical areas of Pakistan and circumstances. Authors designed an “Energy Analytical Data Mart” to store the historical and continuously growing data. In this case study we have presented two approaches of forecasting the level of natural gas energy production in Pakistan using the artificial intelligence field neural network.