A hybrid grid-based many-objective optimisation algorithm for software defect prediction. (15th January 2021)

• Record Type: Journal Article
• Authors: Wang, Junyan
• Is Part Of: International journal of computing science and mathematics. Volume 12:Number 4(2020)
• Page Start: 374
• Page End: 384
• Publication Date: 2021-01-15
• Subjects: software defect prediction problem -- the probability of detection -- false alarm rate -- adaptive dominant region operator -- convergence -- diversity -- many-objective optimisation; Mathematics -- Periodicals; Computer science -- Periodicals; Mathematics -- Data processing -- Periodicals; 510.285
• Languages: English
• ISSNs: 1752-5055

Abstract

How to apply limited test resources to detect error module is one of the challenges of software defect prediction problem. To solve the problem, a many-objective software defect prediction model is proposed by considering the probability of detection and false alarm rate, the Balance value and F-measure as defect prediction objectives. At the same time, a hybrid grid-based many-objective optimisation algorithm is designed to solve the model. In the designed algorithm, the adaptive dominant region operator is introduced into the grid-based many-objective optimisation algorithm to improve the performance of algorithm in balancing dynamically the convergence and diversity of population. The simulation results show that the proposed algorithm has better performance in solving many-objective the software defect prediction problem.