Abstract
Classification builds a model based on historical data (training data set). Once the model is built, it is used to predict the class for a new instance. Many methods have been proposed to solve the classification problem (referred to as classifiers); one of the most popular and best classifiers proposed so far is the decision tree classifier. Multiple trees can be generated from the same dataset, all the trees yields the same outcome for a given new instance to be classified. The possible trees for a dataset vary in their size where the size of the tree depends on the sequence in which the dataset attributes is used to build the tree. However, we prefer the minimum tree because the minimum tree needs the shortest time to figure out the outcome of the model. One of the best algorithms that have been proposed to find the sequence that yield the minimum tree if used is the entropy algorithm. We proposed in this article a new algorithm (enhanced entropy algorithm) that reduces complexity and execution time of the original entropy algorithm and at the same time yields the same sequence that can be found by applying entropy algorithm.