Saturday, March 1, 2014

JMetals

1. How to change the initialize population (in algorithms)
2. How to encode in int[] (solution type)
3. How to make it evaluate fitness function (in problem, evaluate, set objective value for a solution(gene))
4. What is the use of evaluateConstraints (in problem, evaluate, set #constraintsViolated for a solution(gene))
5. How to sort the population (there is a sort function in algorithm.execute())
6. How to change the evaluation of crossover and mutation (is it working on clone)? (Yes)
7. What I should do about for quality indicator
8. The objective functions in the problem are to be minimized (the smaller the better).


Main(Problem, Algorithm, Operator, parameters)
Problem (the problem to be solve, #obj,#constraint,#var) ->SolutionType(Types that is used by the genes) -> CreateVariables(Create contents in the genes for initial population)

Algorithm (e.g., GA, execute()) -> create initial population (SolutionSet(Population) -> Solution (Gene Instance))


For IBEA
Spread, etc. are just quality indicators that it is used to judged the multiobjective optimization, it is not used in the process. The indicator used in the process is the hypervolume indicator.

No comments:

Post a Comment