Selection of Machining Parameters in Ultrasonic Machining Process Using CART Algorithm

Data mining, also referred as ‘knowledge mining from data’, is the process in which useful information is extracted from a data and is transformed into an understandable pattern or structure. In this paper, one of the major data mining techniques known as classification method is applied to identify the most important machining parameters of ultrasonic machining of WC-Co composite material. The decision tree is developed by classification and regression tree algorithm (CART) to analyse the most predominant process parameters affecting the responses. The results of decision tree presented the effects and the major contribution of various machining parameters on material removal rate and tool wear rate. Analysis of variance is also applied to validate the outcomes of considered data mining technique.