Title: Evaluation model of college students' online learning level difference based on support vector machine

Authors: Kunzhe Liu; Xiqiang Ge; Jiefeng Wang

Addresses: College of Education, Fuyang Normal University, Fuyang, 236037, China ' College of Education, Fuyang Normal University, Fuyang, 236037, China ' College of Education, Fuyang Normal University, Fuyang, 236037, China

Abstract: There are some problems in the evaluation of online learning level differences among college students, such as low accuracy and long evaluation time. This paper proposes a new research method by extracting online learning behaviour characteristics to determine behaviour evaluation indexes and evaluation standards. The depth of the residual neural network method is used to remove interference index. Using the Lagrange multiplier method, the evaluation problem is transformed into a dual problem. The nonlinear transformation causes us to strive for the optimal classification plane. Difference in evaluation data was accessed by using big data technology, through the normalised processing the data, in order to obtain the optimal classification function. Finally, the final evaluation results are obtained, and the online learning evaluation level differences are completed. Through comparison, the accuracy of this method is 97%, and the time cost of assessment is always less than 9.5 ms.

Keywords: support vector machine; online learning; difference evaluation; behaviour characteristics.

DOI: 10.1504/IJCEELL.2023.127849

International Journal of Continuing Engineering Education and Life-Long Learning, 2023 Vol.33 No.1, pp.23 - 36

Received: 17 Nov 2020
Accepted: 18 Jan 2021
Published online: 20 Dec 2022 *