PROJECT TITLE :
Behavior of Multiobjective Evolutionary Algorithms on Many-Objective Knapsack Problems
We tend to examine the behavior of three categories of evolutionary multiobjective optimization (EMO) algorithms on many-objective knapsack issues. They are Pareto dominance-primarily based, scalarizing perform-based mostly, and hypervolume-based mostly algorithms. NSGA-II, MOEA/D, SMS-EMOA, and HypE are examined using knapsack problems with a pair of-ten objectives. Our test issues are generated by randomly specifying coefficients (i.e., profits) in objectives. We additionally generate different test problems by combining 2 objectives to create a dependent or correlated objective. Experimental results on randomly generated several-objective knapsack issues are consistent with well-known performance deterioration of Pareto dominance-based algorithms. That's, NSGA-II is outperformed by the other algorithms. But, it's also shown that NSGA-II outperforms the other algorithms when objectives are highly correlated. MOEA/D shows totally completely different search behavior relying on the choice of a scalarizing perform and its parameter worth. Some MOEA/D variants work very well only on 2-objective issues whereas others work well on several-objective issues with four-10 objectives. We have a tendency to conjointly acquire other fascinating observations like the performance improvement by similar parent recombination and the need of diversity improvement for many-objective knapsack issues.
Did you like this research project?
To get this research project Guidelines, Training and Code... Click Here