Implementation of Genetic Algorithms to Cluster New Students into Their Classes

Abstract

Clustering new students into their classes at random probably make an educational problem, because the smartest student maybe clustered in a same class with the most stupid one. To avoid this problem, we can use sorting-score method which cluster new students based on their achievements. First, we sort the average of their scores, and then make the clusters(classes) base on its. This method is not so worse than the first one, but only the smartest class and the most stupid class that have a low gap. There are high gaps in the middle classes. This research tries to explore Genetic Algorithm (GA) to solve this problem. Experimental studies show that performance of GA is better than sorting-score method. The gap of intelligence in classes clustered by GA are relatively same each other. GA can reduce maximum gap in class that clustered by sorting-score method.
Keywords: cluster, Genetic Algorithm, similarity, student.