Title: Big data classification of learning behaviour based on data reduction and ensemble learning
Authors: Taotao Wang; Xiaoxuan Wu
Addresses: Department of Information Engineering College, Jiangxi University of Technology, JiangXi 330098, China ' Department of General Education, Guangxi Vocational College of Water Resources and Electric Power, Nanning 530023, China
Abstract: In order to overcome the problems of low classification accuracy, long time, and high missing ratio of traditional methods, a big data classification method of learning behaviour based on data reduction and ensemble learning was proposed. By cleaning and transforming the big data of learning behaviour and discretising the attributes of big data of learning behaviour, the data reduction algorithm is used to simplify the attributes of big data of learning behaviour. The ensemble learning method is used to linearly combine several weak classifiers, and the ensemble classifier is trained according to Choquet integral. The trained classifier is used to classify the big data of learning behaviour after simplified processing. The experimental results show that when the amount of big data on learning behaviour reaches 5,000 GB, the average classification accuracy of the proposed method is 92%, the classification time is 29 s, and the failure rate of classification is 0.32%.
Keywords: data reduction; ensemble learning; rough set theory; big data of learning behaviour; big data classification.
DOI: 10.1504/IJCEELL.2023.132418
International Journal of Continuing Engineering Education and Life-Long Learning, 2023 Vol.33 No.4/5, pp.496 - 510
Received: 08 May 2021
Accepted: 09 Aug 2021
Published online: 19 Jul 2023 *