Abstract
Predicting the future is always a quest of mankind and thus Supervised Learning (predictive modeling, machine learning) is one of the most rapidly used techniques of Data Mining. Finding out patterns and measuring the accuracies of the foretell are very hot research areas these days. In this research paper, we have introduced a new method to generate a final optimal and accurate classifier from several interim classification trees for various samples of same dataset. This method saves plenty of time of passing test data from several trees because instead we will have an ultimate classifier from the merger of the interim trees, with the help of information gain theory. In this paper, the method is applied on the Drug Data used in SPSS Clementine demonstration. Above all, the proposed method is quite simple and easy to understand as well as easy to implement in practical environment.