Zhang Ling. Research on Prediction of massive open online course’s Academic Achievement of Computer Major Based on BMA and Improved K- Nearest Neighbor Model. Dynamic Systems and Applications 29 (2020) No. 4, 1499 – 1512

https://doi.org/10.46719/dsa202029436

ABSTRACT.
With the rich digital resources and the support and assistance of various learning support systems, online learning has become a widely used learning method. In this paper, the daily learning data generated by computer majors on the platform are taken as samples, and the five most representative influencing factors are synthesized. The samples are classified by the improved K- nearest neighbor model, and the online learning performance prediction data based on BMA is obtained. Using the online learning behavior data of computer major students as experimental data, by comparing the prediction results of computer major students’ massive open online course, different algorithm models are used for different courses, and the accuracy rate is analyzed and a conclusion is drawn. Experimental results show that the k- nearest neighbor algorithm optimized by BMA makes the prediction accuracy higher.

Key words: BMA; Improved k-nearest neighbor model; Studying in massive open online course; Achievement prediction