Abstract—The latest coal mine policy of China puts forward new requirements for safe mining in the coalmine. This paper proposes a BP neural network ensemble learning model based on additional momentum term combined with Bagging algorithm for the prediction of coal and gas outburst. The additional momentum term approach is used to optimize the weights and thresholds of BP neural network during the training process. In order to enhance the prediction ability of BP neural network, the Bagging algorithm is adopted to build ensemble neural network. Experiments show that as compared with the standard BP neural network, the ensemble learning model can improve the prediction accuracy of 8%-10% points in the prediction of coal and gas outburst and the prediction accuracy is over 95%.
Index Terms—BP neural network, Bagging algorithm, ensemble learning model, additional momentum terms, coal and gas outburst prediction.
Jianxing Liao, Junwei Lv, Yuhan Shao, and Peipei Li are with Hefei University of Technology, Hefei 230601, China (e-mail: ljx4805@163.com, junweilv@hfut.edu.cn, yuhan_shao@yeah.net, peipeili@hfut.edu.cn).
Xuegang Hu is with the Research Association of Computer Education in colleges and University of Anhui Province, and School of Computer Science and Information Engineering, Hefei University of Technology, Hefei 230601, China (e-mail: jsjxhuxg@hfut.com).
Cite: Jianxing Liao, Junwei Lv, Yuhan Shao, Peipei Li, and Xuegang Hu, "Application of BP Neural Network Ensemble Model Based on Bagging Algorithm," International Journal of Machine Learning and Computing vol. 9, no. 2, pp. 121-128, 2019.