Journal of Emerging Trends in Engineering and Applied Sciences: Research Areas

Abstract:
The growing interest in the study of multicriteria scheduling problems has led to the development of a number of approximations algorithms for the problem. This has resulted in the need to assess the performance of these approximation algorithms. To this end, many metrics have been proposed. However, there is concern that some of these metrics may produce conflicting results. This then makes comparison of such approximation algorithms difficult. In this paper, a sample of these metrics were tested on a large number of non-dominated (NDS) sets in order to determine whether there is agreement in terms of their assessment of the best non-dominated set for each instance of the problem. Experimental results show that, to a large extent, there is agreement among the metrics when they are combined in assessing the performances of the non-dominated sets rather than when they are used individually.

Keywords: multi-objectives scheduling, agreement, metrics, performance, jobs, machines

Download full paper