ABSTRACT
In data-mining, many algorithms can’t process continuous value, so, discretization is an important method in data-mining application. Discretization describes converting some continuous data into some discrete values. In our paper firstly, we will try to implement three discretization techniques on a dataset, namely, Equal Width, Global Equal Width and Equal Frequency. Secondly we will apply two causality finding techniques namely, Mutual Information and Transfer Entropy and compare the performance among those discretization techniques.
