Analysis on the determination of trainees based on an emerging machine learning LightGBM

Abstract:
The rise of social military, military training, improve the pertinence of training selection is the current concern. In order to scientifically determine whether the personnel need to be trained again, GBDT algorithm, XGBoost and LightGBM are proposed. At the same time, the obtained data set 8: 2 is divided into training set and test set, and the prediction model of the personnel to be trained is established based on the selected characteristics. The prediction performance of the three algorithms is comprehensively compared on the real data. The results show that LightGBM algorithm is more accurate in predicting whether a person needs to be trained again. This provides a training evaluation idea for managers in the training field to provide prospect.