Teaching Evaluation Research of Professional Courses Based on Multinomial Random Forest With Improved Grassberger Entropy

Abstract

This study proposes a multinomial random forest algorithm based on improved Grassberger entropy to evaluate the teaching effect of professional courses. To reduce calculation deviation, the proposed algorithm uses improved Grassberger entropy to calculate information gain and selects an optimal split attribute of a split node to train a multinomial random forest classifier. Two multinomial distributions based on information gain are used instead of simple randomness to realize the random selection of a split feature and split value to increase classification consistency. The proposed algorithm is used to construct a course teaching effect prediction model to analyze the course learning effect on students. This study also examines and determines the evaluation indicators that should be included in the curriculum teaching effect evaluation system and constructs a student-oriented evaluation indicator system. The experimental results indicate that the proposed model can accurately identify student grades of different evaluation results obtained from evaluation index data containing noise. Also, the analysis results indicate the importance of evaluating each indicator characteristic for effectively improving the teaching effect.

Keywords

• Evaluation indicator
• Grassberger entropy
• multinomial random forest algorithm
• teaching quality assessment