Application of ensemble learning in evaluating the effectiveness of military training
Weiwei Zhang, Minglong Lu, Yaoqiang Liu, Yuan Mei, Xiao Jing Cai, Junpeng Zhao · 2023
In order to solve the problem of subjective adjustment of evaluation index weights in the Analytic Hierarchy Process (AHP) for military training effectiveness evaluation, and the inability to accumulate historical evaluation experience, which leads to unscientific comparison of effectiveness evaluation results between different units, this study proposes an integrated learning model applied in the field of military training effectiveness evaluation. This model is based on the existing Analytic Hierarchy Process and constructs an indicator system based on historical evaluation data of each unit, Train multiple sub models separately, and through the designed model fusion device, fuse the multiple sub models into the military training effectiveness evaluation model proposed in this paper. Using a certain training dataset as the data source, the samples were divided into training and testing sets in a ratio of 8:2. After training, the proposed model achieved an accuracy of 98.98% in the testing set, with an average absolute error of 1.08%. This model effectively avoids subjective evaluation and can provide scientifically comparable evaluation results.