Abstract | ||
---|---|---|
Concept-based MOEAs are tailored MOEAs that aim at solving problems with a-priori defined subsets of solutions that represent conceptual solutions. In general, the concepts' subsets may be associated with different search spaces and the related mapping into a mutual objective space could have different characteristics from one concept to the other. Of a particular interest are characteristics that may cause premature convergence due to local Pareto-optimal sets within at least one of the concept subsets. First, the known ε-MOEA is tailored to cope with the aforementioned problem. Next, the performance of the new algorithm is compared with C1-NSGA-II. Concept-based test cases are devised and studied. In addition to demonstrating the significance of premature convergence in concept-based problems, the presented comparison suggests that the proposed tailored MOEA should be preferred over C1-NSGA-II. Suggestions for future work are also included. |
Year | DOI | Venue |
---|---|---|
2012 | 10.1007/978-3-642-32964-7_13 | PPSN (2) |
Keywords | Field | DocType |
future work,concept-based moeas,aforementioned problem,premature convergence,conceptual solution,different characteristic,concept subsets,concept-based test case,concept-based problem,different search space | Mathematical optimization,Premature convergence,Computer science,Test case,Concept selection | Conference |
Citations | PageRank | References |
2 | 0.43 | 7 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Amiram Moshaiov | 1 | 62 | 11.10 |
Yafit Snir | 2 | 2 | 0.43 |