Hybrid Test Case Optimization Approach Using Genetic Algorithm With Adaptive Neuro Fuzzy Inference System for Regression Testing

Abstract

Abstract In an agile environment, regression testing is inevitable because it aims to attain software with better quality. Regression testing identifies whether an accurate result can be obtained for the corresponding input submitted. The test cases, hence, have an important role to play in the error-identification process of an application. An algorithm has been proposed in this paper that prioritizes the test cases based on the rate of fault detection and impact of faults. Artificial intelligence techniques have been incorporated with optimizing algorithms to reduce the overall number of test cases using minimization, selection, and prioritization collectively. Fuzzy K-means (FKM) clustering algorithm is used to cluster the test cases initially. The clustered test cases are further scaled down using the genetic algorithm (GA) combined with the adaptive neuro fuzzy inference system (ANFIS) called “Hybrid G-ANFIS.” The size of the test suite is reduced because only the optimal test cases are selected for further optimization by the ANFIS, from an already clustered test suite. This is done by fuzzy logic principles that select only the test cases that are needed for validating the changes in the software. Along with this, the test cases that have the ability to find faults by covering maximum code in a minimum time frame are also chosen. Optimization achieves a better outcome because it is done repetitively both by clustering and optimization algorithms continuously, which results in reducing the test cases considerably in the regression test suite. The proposed research work is evaluated in terms of performance measures, namely fault detection ratio, fault coverage, statement coverage, and an average percentage of faults detected (APFD) chart from which it is clear that a better optimization of the regression testing estimation process can be done.