On the Performance Assessment and Comparison of Stochastic Multiobjective Optimizers
PPSN IV Proceedings of the 4th International Conference on Parallel Problem Solving from Nature
Inferential Performance Assessment of Stochastic Optimisers and the Attainment Function
EMO '01 Proceedings of the First International Conference on Evolutionary Multi-Criterion Optimization
ISDA '05 Proceedings of the 5th International Conference on Intelligent Systems Design and Applications
Design and analysis of stochastic local search for the multiobjective traveling salesman problem
Computers and Operations Research
EA'09 Proceedings of the 9th international conference on Artificial evolution
Adaptive "Anytime" two-phase local search
LION'10 Proceedings of the 4th international conference on Learning and intelligent optimization
EMO'05 Proceedings of the Third international conference on Evolutionary Multi-Criterion Optimization
Performance assessment of multiobjective optimizers: an analysis and review
IEEE Transactions on Evolutionary Computation
IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans
Tool sequence optimisation using preferential multi-objective search
Proceedings of the 15th annual conference companion on Genetic and evolutionary computation
| Hi-index | 0.00 |
A fundamentally different approach to the quality assessment of multi-objective SLS algorithms derives from the concept of attainment function. The attainment function extends the scalar concepts of mean and variance to random sets. The attainment function theory may completely characterize the statistical distribution of solutions in the objective space in terms of location, spread and mutual dependence. Moreover, statistical testing and inference are possible. However, the use of attainment functions is still rather limited in practice. We present here two practical applications of the first-order attainment function for analysing the output of SLS algorithms for bi-objective optimization problems. Programs implementing the techniques presented here are also available. Later, we discuss what would be necessary to extend this work for more than two objectives and for other types of analysis.