An approach based on decision tree for analysis of behavior with combined cycle power plant

Abstract

This paper presents about Combined Cycle Power Plant (CCPP) and decision tree. CCPP considered as the best effective power suppliers to the large temperature incline between its gas turbine passage and the environment or the cooling process, and to help of their engineers, who are able to optimally venture the present temperature level. Moreover, in this paper we did comparison of four types of decision tree algorithms like Decision Stump, Hoeffding Tree, logistic model trees (LMT) and J48. Based on these algorithms we analyzed the behavior of Combined Cycle Power Plant (CCPP), particularly its temperature. The temperature is target variable among other variables. This is why temperature was divided three classes. Theoretical analysis and experimental results have shown that the J48 algorithm is the best algorithm which predict attributes of given instances precisely among other three algorithms. Based on our findings, J48 algorithm can predict precisely the temperature of Combined Cycle Power Plant and predict 97.1676% cases correctly.