Evolutionary Algorithms

Question 1

Situation when we lack detailed information about the underlying optimization problem.

Answer 1

Black box

Question 2

Family of robust and universally applicable optimization methods, their flexibility enables usage within black-box setting.

Answer 2

Evolutionary algorithms

Question 3

The most simple version of an evolutionary algorithm

Answer 3

Evolutionary strategy

Question 4

The two levels of existence in the representation of an evolutionary algorithm.

Answer 4

Genotype and phenotype

Question 5

Possible solutions in the original problem context

Answer 5

Phenotype

Question 6

Encoding of the phenotype in the context of an EA

Answer 6

Genotype

Question 7

Function that represents the optimization target in a EA.

Answer 7

Fitness function

Question 8

Single real-valued number to which each phenotype is mapped to.

Answer 8

Fitness value

Question 9

The current set of candidate solutions/individuals, can contain multiple copies of the same individual.

Answer 9

Population

Question 10

(𝜇 + 𝜆)-selection which selects, the best 𝜇 individuals out of 𝑃𝜇 ∪ 𝑃𝜆, enables ___.

Answer 10

Elitism

Question 11

(𝜇, 𝜆)-selection select the best 𝜇 individuals only out of 𝑃𝜆, enables ___.

Answer 11

Random walks

Question 12

They are responsible for the generation of new candidate solutions.

Answer 12

Variation operators

Question 13

The two types of variation operators.

Answer 13

Mutation and recombination

Question 14

An individual that survives a long time in an evolutionary algorithm is called a(n):

Answer 14

Elite

Question 15

True or false? Evolutionary algorithms are a suitable method for dealing with black-box optimization problems.

Answer 15

True

Question 16

True or false? Mutation in evolutionary algorithms is applied to get better exploitation of the search space.

Answer 16

False. Mutation is responsible for the exploration of the search space

Question 17

True or false? in (μ+λ) -selection, individuals from the parent generation can survive and become part of the next generation.

Answer 17

True

Question 18

True or false? The landscape of the feature space, i.e., joint space of search space and fitness function, is a priori usually unknown.

Answer 18

True

Question 19

Which variation operator is responsible for exploring the search space?

Answer 19

Mutation