engleski

Data Mining and Statistical Analyses for High Education Improvement

Knowledge creation is essential for process improvement in every type of environment. This is also the case in lecturing and grading students at universities. In this paper a visual representation of data is combined with statistical analysis in order to be used for analyzing objectivity of student testing. Predictive data mining is used for timely recognition of students who require additional attention. Regression analysis and decision trees are methods used for this student selection. Comparison of these two methods, as well as a simple method of score border is also given.

education; testing objectivity; statistical analyses; predictive data mining

nije evidentirano