Application of XGboost Algorithm in Bearing Fault Diagnosis

[摘要] This paper applies the XGboost(eXtreme Gradient Boosting) algorithm to the fault diagnosis of rolling bearing. XGboost is the realization of GBDT(gradient boosting decision tree). Generally speaking, the realization of GBDT(gradient boosting decision tree) is slow. XGBoost is characterized by fast computation and good performance of the model. At the end of this paper, we compare with other tree algorithms, and the results show that the XGboost algorithm is superior to other algorithms in accuracy and time.

[发布机构] School of Electrical Engineering, Shanghai Dianji University, Shanghai, China^1

[效力级别] 能源学 [学科分类] 材料科学（综合）

[关键词] Bearing fault diagnosis;Fast computation;Gradient boosting;Rolling bearings;Tree algorithms