Search. Read. Cite.

Easy to search. Easy to read. Easy to cite with credible sources.

Information Technology Journal

Year: 2011  |  Volume: 10  |  Issue: 8  |  Page No.: 1508 - 1517

Applying Machine Learning Methods to the Shooting Accuracy Prediction: A Case Study of T-75 Pistol Shooting

S. Deng, D.M. Liu and S.L. Hsieh

Abstract

In this study, we investigate the quantitative correlation between human factors and the shooting accuracy of the T75 assault pistol. We carry out live pistol firing. Eleven shooters are selected to fire 30 rounds each. We obtain 330 data successfully. Interest factors are assigned to three human factors, whereas performance evaluation methods are assigned to three parameters of impact points; these experiment datasets are measured using an I-CubeX glove force sensor system. Three prediction models (for shooting score, shooting precision and shooting trueness) are established by using a Least Squares Support Vector Machine (LS-SVM), Back-Propagation Neural Network (BPNN) and Response Surface Methodology (RSM); the results of present study indicate that the model developed by using an LS-SVM exhibits excellent prediction ability. The force of the shooter’s right index finger abdomen for pulling the pistol trigger and the force of the shooter’s left palm for gripping the pistol significantly influence shooting performance. An inexperienced shooter can use the results of this study as a reference for improving his or her shooting skills. Furthermore, this study will greatly assist efforts to upgrade pistol design and performance.

Cited References Fulltext