October 14, 2021 (same as October 2021, 2019)
1. The following picture shows the maximum and average fitness curves of an evolving population. What
can we infer regarding the population diversity at generation n?
A Nothing
B The first derivative of the diversity curve is zero
C Diversity must be at its maximum
D Diversity must be at its minimum
Note: The population can have different genotypes that all map to the same phenotype.
2. We tackle the n-queens problem with a GA using a bitstring representation where 1 (0)
denotes the presence (absence) of a queen on a square. What is the dimensionality of the
search space?
A 2n
B n!
C n2
Dn
3. We want to optimise the function f (x, y) = x + y with Differential Evolution. Consider
the following population of 6 individuals:
The first step in creating the next generation is the creation of a mutant vector population. What is mutant
vector ⊽4 if the base vector ā4 is individual 5, the difference vector 1 is defined by b̅4 = individual 1 and c̅4 =
individual 2, and the scaling factor is F = 0.5?
A ⊽4 = 〈0.2, 0.5〉
B ⊽4 = 〈0.25, 0.9〉
C ⊽4 = 〈0.35, 0.9〉
D ⊽4 = 〈0.4, 1.0〉
4. What is parameter tuning?
A Parameter tuning is adjusting parameters of the evolutionary algorithm before a run
B Parameter tuning is adjusting parameters of the evolutionary algorithm during a run
C Parameter tuning is adjusting parameters of the evolutionary algorithm during a run based on time
D Parameter tuning is adjusting parameters of the evolutionary algorithm by coding them in
the genome
, 5. Determine the truth of statements I, II, and III and select the correct answer from A, B,
C, or D.
I Mutation is a unary reproduction operator.
II Crossover is a unary reproduction operator.
III Recombination is an n-ary reproduction operator (n ≥2).
A I and II are true, while III is false
B I is true, II and III are false
C I and III are true, II is false
D All three statements are true
October 15, 2018
1. Which of the following statements is true?
A Evolution Strategies do not use genotypes, they work directly on the phenotypes
B Evolution Strategies do not use phenotypes, they work directly on the genotypes
C In Evolution Strategies the genotype space and the phenotype space are identical
D In Evolution Strategies the phenotype space and the fitness space are identical
2. Consider the following illustration of two parents and a child in a 2-dimensional real valued search
space. Point Z is the child of X and Y obtained through
A Single arithmetic crossover
B Whole arithmetic crossover
C Blend crossover
D Simple arithmetic crossover
3. When defining problem complexity (“difficulty”), the definition of class P relies on the characteristics of
A The problem at hand
B The algorithms that can solve the problem at hand
C Both
D None; it concerns the search space.
4. The picture shows that the maximum and average fitness curves touch each otherat generation n, stay
equal for a few generations and split again after generation N. Wha tis the most likely cause of this split at
generation N?
A Cannot tell, could be anything
B A stronger selection scheme was activated
C A softer selection scheme was activated
D A successful mutation took place
1. The following picture shows the maximum and average fitness curves of an evolving population. What
can we infer regarding the population diversity at generation n?
A Nothing
B The first derivative of the diversity curve is zero
C Diversity must be at its maximum
D Diversity must be at its minimum
Note: The population can have different genotypes that all map to the same phenotype.
2. We tackle the n-queens problem with a GA using a bitstring representation where 1 (0)
denotes the presence (absence) of a queen on a square. What is the dimensionality of the
search space?
A 2n
B n!
C n2
Dn
3. We want to optimise the function f (x, y) = x + y with Differential Evolution. Consider
the following population of 6 individuals:
The first step in creating the next generation is the creation of a mutant vector population. What is mutant
vector ⊽4 if the base vector ā4 is individual 5, the difference vector 1 is defined by b̅4 = individual 1 and c̅4 =
individual 2, and the scaling factor is F = 0.5?
A ⊽4 = 〈0.2, 0.5〉
B ⊽4 = 〈0.25, 0.9〉
C ⊽4 = 〈0.35, 0.9〉
D ⊽4 = 〈0.4, 1.0〉
4. What is parameter tuning?
A Parameter tuning is adjusting parameters of the evolutionary algorithm before a run
B Parameter tuning is adjusting parameters of the evolutionary algorithm during a run
C Parameter tuning is adjusting parameters of the evolutionary algorithm during a run based on time
D Parameter tuning is adjusting parameters of the evolutionary algorithm by coding them in
the genome
, 5. Determine the truth of statements I, II, and III and select the correct answer from A, B,
C, or D.
I Mutation is a unary reproduction operator.
II Crossover is a unary reproduction operator.
III Recombination is an n-ary reproduction operator (n ≥2).
A I and II are true, while III is false
B I is true, II and III are false
C I and III are true, II is false
D All three statements are true
October 15, 2018
1. Which of the following statements is true?
A Evolution Strategies do not use genotypes, they work directly on the phenotypes
B Evolution Strategies do not use phenotypes, they work directly on the genotypes
C In Evolution Strategies the genotype space and the phenotype space are identical
D In Evolution Strategies the phenotype space and the fitness space are identical
2. Consider the following illustration of two parents and a child in a 2-dimensional real valued search
space. Point Z is the child of X and Y obtained through
A Single arithmetic crossover
B Whole arithmetic crossover
C Blend crossover
D Simple arithmetic crossover
3. When defining problem complexity (“difficulty”), the definition of class P relies on the characteristics of
A The problem at hand
B The algorithms that can solve the problem at hand
C Both
D None; it concerns the search space.
4. The picture shows that the maximum and average fitness curves touch each otherat generation n, stay
equal for a few generations and split again after generation N. Wha tis the most likely cause of this split at
generation N?
A Cannot tell, could be anything
B A stronger selection scheme was activated
C A softer selection scheme was activated
D A successful mutation took place
