Comparative Study of Binary Classification Algorithms to Analyze the Students' Performance on Virtual Machine

In recent times, the utilization of learning management systems in education has been increased to a great extent. In order to access online contents, students have started to use mobile phones, especially smart phones. Student's online activities generate immeasurable amount of data that cannot be processed by traditional tools and techniques. This has resulted in the invasion of Big Data technologies and tools into education, to process the enormous amount of data involved. The large volume of data collected for many years contains hidden knowledge, which helps to improve the students? performance. The big data collected from institutions are analyzed in a distributed environment in order to maintain the efficiency and reduce the computational complexity. Machine learning algorithms are used to find meaningful patterns that are hidden in the data. In this paper, machine learning algorithms such as Decision Tree, Na?ve Bayes and NBTree are analyzed in both local machine as well as virtual machine on higher secondary school students? data. The data set consists of 17 features and 115328 records. The analysis is based on the parameters such as accuracy, execution time and memory usage. The outcome of our study shows that the NBTree algorithm gives the highest accuracy of 98% with execution time of 0.234 seconds and 49.42 MB of memory usage in local mode. Similarly, the highest accuracy of 99% was achieved with execution time of 0.184 seconds, 45.42 MB of memory usage in virtual mode.