Data mining is the process of discovering useful patterns and relationships from the data. It facilitates making decision in education field. The quantity of data educational database stores is growing rapidly. Higher educational institute can use this database which holds hidden information. This hidden information is helpful in improving student’s performance in their academics. Successful prediction of student’s performance leads to assure and achieve better quality in higher education system. The performance of student is determined by different attributes. In this paper, the classical techniques and the evolutionary data mining technique called grammar based genetic algorithm (GGA) is used to build a model for predicting the performance of undergraduate students by using their past educational and general record. The result of different classifiers is compared to discover the final prediction model. Further, the result of prediction is analyzed and its relationship with different predictors is studied. This kind of analysis facilitates improvement in a specific area of educational system and also helps studying the various general factors responsible for the performance of student. © Research India Publications.