A Comparison of the Efficiency of Data Classification in Learning Factors of Tertiary Level Students in a Virtual Learning Environment
This research presents the result of evaluation and comparison of the efficiency of data classification in learning support factors of tertiary level students in a virtual learning environment. Three primary techniques were compared in this research, including Decision Tree technique, k-NN technique, and Naive Bayes technique. The were 8 attributes with 32,950 datasets of the data which were included student registration and the learning outcome of the students who interactively learned in a virtual learning environment. The data was divided into 2 parts by Split Test method. The first part which was comprised of 26,360 datasets (approximately 80 percent of the total data) was used to generate a model while the second part which was included 6,590 datasets (approximately 20 percent of the total data) was utilized to examine the precision and accuracy of the model. The results revealed that the most efficient technique, used to classify the learning factors of the tertiary students in a virtual learning environment, was k-NN technique (k = 3) with 97.87 percent of precision, 97.90 percent of recall, 97.93 percent of accuracy, and 0.352 F-measure.
Classification; Split Test; Decision Tree; k-NN; naive bayes
- There are currently no refbacks.