Execution Time Forecasting of Automatic Test Case Generation Based on Genetic Algorithm and BP Neural Network

Abstract

The time spent on automatic test case generation is an important parameter in the code defect detection technology. The accurate forecast of the time spent on automatic test case generation is critical to the efficiency of code defect detection. This paper applied the BP neural network to the forecast of the time spent on automatic test case generation is to consider the factors of the number of constraint variables, the number of function calls, the number of constraint expressions and so on as the input unit of BP neural network. The results show that compared with the traditional BP neural network, the BP neural network optimized by genetic algorithm can speed up the convergence rate of the network and improve the forecasted accuracy of the time spent on automatic test case generation.