Machine Learning based Accuracy Prediction Model for Augmented Biofeedback in Precision Shooting

Abstract

Abstract In military, police, security companies, and shooting sports, precision shooting training is of the outmost importance. In order to achieve high accuracy, trainees need to do a lot of training. Consequently, they will consume a great number of rounds (cartridges) and a considerable amount of professional coaches’ time - both could cost much. Our motivation is to reduce the costs and shorten the training time by implementing an augmented biofeedback system based on machine learning techniques. We designed a biofeedback system, which can detect and give feedback about three kinds of errors that regularly occur during precision shooting practice: excessive hand movement error, aiming error and triggering error. The system provides concurrent feedback about the first error and terminal feedback about the last two errors. Machine learning techniques are used for the identification of hand movement errors. A precision shooting accuracy prediction model based on random forest (RF) has been found as the most appropriate. The experimental results show that: (a) the proposed RF model achieves the prediction accuracy of 91.27%, higher than any of the existing reference models, and (b) the hand movement is strongly related to the accuracy of the precision shooting. An appropriate use of our system can lead to the reduced number of rounds used and to the disburdening of coaches as the trainee can learn about the most common mistakes from the system during the training.