An empirical study of test effort estimation based on bat algorithm

Abstract

The process of actually predicting efforts for software testing phase is a complex task. There are many factors affect test effort estimation, including productivity of the test team, strategy chosen for testing, size and complexity of the system, technical factors, expected quality and others. Several studies have been done for developing test effort estimation models but to some extent, most of these models result in erroneous effort estimation. Thus, there is a strong need to optimise the test effort estimation. In this paper, we proposed a model using the meta-heuristic bat algorithm to estimate the test effort. The proposed model is then used to optimise the effort by iteratively improving the solutions. Results show that our estimations is closer to the actual efforts and is thus more accurate than other methods.