chapter 4 Flashcards

1
Q

what is a representation?

A
  1. genotype space
  2. mapping

>>> often representation is used to refer to the genotype space

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

what is representation discontinuity?

A

representation discontinuity:

> difference in phenotypic vs genotypic distance

> e.g. phenotype = positive ints, genotype = bit strings

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

what is the strong duality principle?

A

strong duality:

> small changes in underlying structuve of an object should cause small changes in the objects behavior

> crucial concept in EC

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

binary representation: how does 1 point crossover work?

A

1 point crossover

  1. choose random point on the two parents
  2. split parents at this point
  3. create children by exchanging tails
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

binary representation: whats a problem with one point crossover?

A

positional bias

> keeps genes together that are close to each other

> performance depends on the order the variables occur in the representation

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

binary representation: how does n point crossover work?

A

n point crossover

  1. choose n random crossover points
  2. split along those points
  3. glue parts together alternating between parents

>>> still has some amount of positional bias

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

binary representation: how does uniform crossover work?

A

uniform crossover:

  1. flip coin for each gene of the first child
  2. inverse selection for the second child

>>> inheritance is independent of position

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

crossover is …?

mutation is …?

A

crossover is explorative

mutation is exploitative

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

real value representation: how does uniform mutation work?

A

uniform mutation:

  1. x’ drawn randomly from [lowerbound, upperbound]

>>> analoguous to bit flipping

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

real value representation: how does non-uniform crossover work?

> whats the step size

A

non- uniform crossover:

  1. most common: add random noise to each var separately from gauss

>>> std is step size

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

real value representation: how does self adaptive mutation work?

A

self adaptive mutation:

> step sizes are included in the genome and undergo variation and selection as well

> step is not size by user but co-evolves

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

why is order of mutation important?

A
  1. first mutate sigma, then x
  2. x’is good is f(x’) is good
  3. sigma’ is good if the x’ it created is good

>>> other way round does not work

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

real valued recombination: two types of recombination

A
  1. discrete: each allele value comes from one of its parents with equal prob
  2. intermediate aka arithmetic crossover

> z = alpha*x + (1-alpha)*y

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

arithmetic crossover 3 possible choices for alpha

A
  1. constant > uniform arithmetical crossover
  2. variable > depending on age of the population
  3. random
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

real valued representation: how does simple arithmetic crossover work

A

simple arithmetic crossover:

> pick random gene, after this point mix values

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

real value representation: how does whole arithmetic crossover work?

A

whole arithmetic crossover:

> mix whole genome

>>> most commong

17
Q

which task typically for permutation representation

A

task: rearrange some objects in a certain order

18
Q

permutation representation: whats a problem with mutation?

A

mutation must change at least two values because otherwise results would be invalid

19
Q

permutation representaiton: what are 4 mutation operators?

A
  1. swap
  2. insert

> pick two alleles

> insert second to follow first, shift rest

>>> preserves most of the order

  1. scrample

> pick subset of genes

> randomly rearrange

  1. inversion

> pick two alleles and invert substring between them

20
Q

what are two characteristics of tree representations?

A
  1. tree shaped chromosomes are nonlinear
  2. trees may vary in depth and width
21
Q

most common recombination in trees?

A

exchange two randomly chosen subtrees among the parents

22
Q

what are 3 key characteristics of reproduction operators?

A
  1. must fit the representation
  2. are stochastic
  3. can be distinguished by arity

> unary: mutation

> binary: recombination and crossover

> n-ary: multi parent