Research on Artificial Intelligence-Assisted Software Test Automation Methods

Software testing faces problems such as low automation and difficult reuse of testing methods. The purpose of this paper is to explore software testing automation methods with the aid of artificial intelligence. In this paper, based on the BS algorithm, the RF algorithm is constructed by Bagging integration, the RF algorithm is optimized by reducing the generalization error through the residual function, and the random forest model for software automation detection is constructed. After that, the model is examined and analyzed using automated detection of malicious software samples as a case study. The experimental results show that the accuracy of the Random Forest algorithm after feature selection reaches 98.9%, its prediction time is the least (7 seconds), and the Random Forest algorithm for training is the best. Z software’s optimized RF algorithm model has an accuracy of between 86% and 99.3% when detecting seven malicious types of samples. This paper’s proposed random forest algorithm model based on artificial intelligence assistance is well-suited for automated software testing, and the detection method is feasible.