Random-Weighted Search-Based Multi-objective Optimization Revisited
Weight-based multi-objective optimization requires assigning appropriate weights using a weight strategy to each of the objectives such that an overall optimal solution can be obtained with a search algorithm. Choosing weights using an appropriate weight strategy has a huge impact on the obtained solutions and thus warrants the need to seek the best weight strategy. In this paper, we propose a new weight strategy called Uniformly Distributed Weights (UDW), which generates weights from uniform distribution, while satisfying a set of user-defined constraints among various cost and effectiveness measures. We compare UDW with two commonly used weight strategies, i.e., Fixed Weights (FW) and Randomly-Assigned Weights (RAW), based on five cost/effectiveness measures for an industrial problem of test minimization defined in the context of Video Conferencing System Product Line developed by Cisco Systems. We empirically evaluate the performance of UDW, FW, and RAW in conjunction with four search algorithms ((1+1) Evolutionary Algorithm (EA), Genetic Algorithm, Alternating Variable Method, and Random Search) using the industrial case study and 500 artificial problems of varying complexity. Results show that UDW along with (1+1) EA achieves the best performance among the other combinations of weight strategies and algorithms.
KeywordsUniformly distributed weights multi-objective optimization search algorithms
Unable to display preview. Download preview PDF.
- 1.Konak, A., Coit, D.W., Smith, A.E.: Multi-objective optimization using genetic algorithms: A tutorial. Reliability Engineering & System Safety (91), 992–1007 (2006)Google Scholar
- 5.Harman, M., Mansouri, S.A., Zhang, Y.: Search Based Software Engineering: A Comprehensive Analysis and Review of Trends Techniques and Applications, Technical Report TR-09-03, King College London (2009)Google Scholar
- 6.Wang, S., Ali, S., Gotlieb, A.: Minimizing Test Suites in Software Product Lines Using Weighted-based Genetic Algorithms. In: Proceedings of the Genetic and Evolutionary Computation Conference (GECCO), pp. 1493–1500 (2013)Google Scholar
- 8.Cisco Systems TelePresence codec c90 (2010)Google Scholar
- 10.Wang, S., Ali, S., Yue, T., Liaaen, M.: Using Feature Model to Support Model-Based Testing of Product Lines: An Industrial Case Study. In: Proceedings of International Conference of Software Quality (QSIC), pp. 75–84 (2013)Google Scholar
- 11.Arcuri, A., Briand, L.C.: A Practical Guide for Using Statistical Tests to Assess Randomized Algorithms in Software Engineering. In: Proceedings of the International Conference on Software Engineering, pp. 21–28 (2011)Google Scholar
- 12.Sheskin, D.J.: Handbook of Parametric and Nonparametric Statistical Procedures (2003)Google Scholar
- 15.Walcott, K.R., Soffa, M.L., Kapfhammer, G.M., Roos, R.S.: Time-Aware Test Suite Prioritization. In: Proceedings of the International Symposium on Software Testing and Analysis, pp. 1–12 (2006)Google Scholar
- 16.Harman, M.: Making the Case for MORTO: Multi Objective Regression Test Optimization. In: Proceedings of the International Conference on Software Testing, pp. 111–114 (2011)Google Scholar
- 17.Smith, N.A., Tromble, R.W.: Sampling Uniformly from the Unit Simplex. Technical Report. Johns Hopkins UniversityGoogle Scholar
- 18.Wang, S., Ali, S., Gotlieb, A.: Random-Weighted Search-Based Multi-Objective Test Suite Optimization Revisited. Technical Report 2013-01 (2013), https://www.simula.no/publications/TR2013-01