Title: A particle swarm optimisation based recurrent neural network method for predicting the trend of Chinese engineering education reform

Authors: Xianjie Peng; Xiaonan Sun; Jiangtao Xu

Addresses: Department of Human Resources, Tongji University, Shanghai, 200092, China ' Institute of Vocational Education, Tongji University, Shanghai, 200092, China ' Hangzhou CITIC Senior Living CORP., Ltd., Zhejiang, Hangzhou, 310002, China

Abstract: Investigating the trends of education reform is important for making plans for future education reform. However, the existing studies mainly adopt simple clustering method to find the relationships among current reform directions. Targeting this problem, this paper first proposes a novel particle swarm optimisation (PSO) with enhanced diversity preservation; second, a PSO-based recurrent neural network (RNN) method is proposed to predict the trend of the engineering education reform in China, where the length and the weights of the RNN are evolved by the proposed PSO method to improve RNN's performance. In the experiment, a dataset of the effort made in Chinese engineering education reform is built and adopted to train the proposed method, and the testing results show that the proposed method outperforms the basic recurrent neural network in prediction, where the testing error of the proposed method is improved with 6.4 compared with the basic RNN.

Keywords: engineering education reform; time series; recurrent neural networks; particle swarm optimisation; diversity preservation.

DOI: 10.1504/IJCSM.2023.131631

International Journal of Computing Science and Mathematics, 2023 Vol.17 No.4, pp.342 - 352

Received: 26 Mar 2022
Accepted: 06 Jul 2022
Published online: 21 Jun 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article