Analysis of the Damage Effect of Direct Aiming Attack on WarGame Replay Data Based on Machine Learning

Through the massive sequence data in the extensive replay data of wargame, logistic regression is one important algorithm of machine learning algorithms used to analyze the damage effects of direct aiming attack in the replay data, the damage effects are generated by various types of attack operators, different types of weapons, and direct aiming attack on different types of target operators. It is useful to train a model through logistic regression algorithm to predict the damage effect, and compare with multiple evaluation indicators by using of dummy features or not. Finally, in future competition, facing the constantly changing Wargame situation, it is crucial to making the decision with the best damage effect. In this paper we use the replay data of the 2023 Wargame competition, firstly we extract the judgement data of direct aiming attack from the massive sequence data of a large number of replay files and build the dataset. Then, study the dataset through visual data analysis and perform feature engineering processing. Finally, compare with multiple evaluation indicators by using of dummy features or not, and obtain a better direct aiming attack damage effect model.