A Modeling of an Intelligent System for Learning Result Prediction to Reduce Drop-Out of Undergraduate Students

Abstract

The objectives of this research were to; 1) analyze the factors of an intelligent system for learning result prediction to reduce drop-out of undergraduate students. 2) construct a modeling of an intelligent system for learning result prediction to reduce drop-out of undergraduate students. The samples were 141 undergraduate students who study English Education program in Academic year 2012-2014 at Kanchanaburi Rajabhat University by purposive sampling. The research results were as follows 1) the factors analysis was based on the attribute weight indexing technique using the Information Gain method. The learning results prediction factors had 14 factors, for example, mean of GPA from semester 1 to 5 and learning results about 9 subjects, 2) constructing a modeling of an intelligent system for learning result prediction to reduce drop-out of undergraduate students by measuring the quality with Cross-validation Test; 10-fold cross-validation and Naïve Bayes technique, the highest accuracy index is 84.33 percent, and followed by the creation of student’s learning result prediction by using Decision Tree technique, 73.86 percent of accuracy index.