Term
| Are genetic algorithms optimal/complete/exhaustive? Why or why not? |
|
Definition
| They are not optimal, complete, or exhaustive. It may never find a solution. It can't be complete or optimal because it isn't exhaustive. |
|
|
Term
| What are the main steps in genetic algorithms? |
|
Definition
Selection
Recombination
Mutation |
|
|
Term
| When are genetic algorithms useful? |
|
Definition
| When time is limited
When progressively better solutions are involved
Can be applied to large state spaces
When there is limited CPU or Memory |
|
|
Term
| When are genetic algorithms a poor choice? |
|
Definition
| When the solution needs to be predictable
When a path needs to be found |
|
|
Term
|
Definition
| level of chromosome's optimality |
|
|
Term
| What is mutation and how does it work? |
|
Definition
| Mutation - Introduces a new gene into the gene pool by putting a random gene in a random spot |
|
|
Term
| What types of selection can be used, and what are the differences? |
|
Definition
| Ranked - only the fittest reproduce
Proportional - Random, but the higher the fitness, the higher the chance of reproduction
Tournament - random genes are selected, two remaining genes reproduce
Random - completely fucking random |
|
|
Term
|
Definition
|
|
Term
|
Definition
| attribute or feature of a solution |
|
|
Term
| What is crossover and how does it work? |
|
Definition
| It is a recombination method where you mix the genes of the parents to create children.
Find a random spot and swap the genes either before or after that spot with the other child |
|
|
