chapter 9 Flashcards
1
Q
4 perspectives on goals
A
- design perspective
> find a very good solution at least once
- production perspective
> find a good solution at almost every run
- publication perspective
> must meet scientific standards
- application perspective
> good enough is good enough
2
Q
2 basic rules of experimentation
A
- EAs are stochastic
> never draw conclusions based on a single run
- EA experimentation is about comparison
> use same amount of resources
> try different comp. limits
> use same performance measure
3
Q
3 performance measures
AES
SR
MBF
A
AES = average number of evaluations to solution
SR = success rate, % of runs finding solutions
MBF = mean best fitness
4
Q
what is the problem with hidden labour
A
hidden labour: heuristic mutation operators might result in bias when comparing numbers of evaluations nbetween algorithms
5
Q
what are 3 general performance measures?
A
- robustness
- speed
- solution quality