An ensemble approach to multi-objective evolutionary algorithm
Multi-objective optimization refers to the procedure of obtaining a set of feasible solution for multiple objective functions. Based on the no free lunch (NFL) theorem, an optimization technique would never exceed all other optimization techniques on every type of optimization problem. Ensemble appr...
Saved in:
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Theses and Dissertations |
| Language: | English |
| Published: |
2019
|
| Subjects: | |
| Online Access: | http://hdl.handle.net/10356/78420 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Nanyang Technological University |
| Language: | English |
| Summary: | Multi-objective optimization refers to the procedure of obtaining a set of feasible solution for multiple objective functions. Based on the no free lunch (NFL) theorem, an optimization technique would never exceed all other optimization techniques on every type of optimization problem. Ensemble approach is one method to improve the performance of the multi-objective algorithm. This method is combining two or more multi-objective algorithms to get the benefit of each individual algorithm.
An ensemble of multi-objective optimization with three multi-objective optimization algorithms (MOEA/D, NSGA-III, SMODE) has been implemented on the multi-objective benchmark test function (a set of many and multi-objective bound constrained benchmark problems). The ensemble method has the best performance in comparison to its former individual algorithm. The results of the simulation show the ensemble has better Pareto-optimal front based on the convergence, diversity and quantitative performance. |
|---|