Selection

Q: How deeply does Optimization Algorithms: AI techniques for design, planning, and control problems cover selection?

Works through fitness-proportionate probabilities and compares selection methods, emphasizing selective pressure and premature convergence rather than formal proofs.

As explained in Optimization Algorithms: AI techniques for design, planning, and control problems

A genetic-algorithm procedure that chooses individuals from a population to serve as parents for reproduction.

Selection decides which candidate solutions get to reproduce and which remain in a genetic algorithm’s population. Parent selection chooses individuals to produce offspring through crossover and mutation. Survivor selection chooses the individuals that form the next population. Keeping these decisions separate helps clarify why a promising parent is not necessarily guaranteed a place in the next generation. section 7.2.3 section 7.3.5

Start with a population whose candidates have fitness values. A selection method favors promising solutions, but how strongly it favors them matters. High selective pressure can reduce diversity quickly and cause premature convergence on a suboptimal solution. Preserve enough opportunity to explore alternatives rather than repeatedly choosing only the current best. section 7.2.3 section 7.3.3

Parent-selection approaches illustrate different ways to favor fitness:

Elitism selects the fittest individuals for reproduction and can carry top performers into the next generation. Preserving more of them reduces the succeeding population’s diversity. section 7.3.3
Fitness-proportionate selection assigns each individual a selection probability proportional to its fitness, then samples randomly. Roulette-wheel selection is one implementation. section 7.3.3

Selecting ticket-pricing candidates

Five candidate pricing solutions have a combined fitness of 329,740. The strongest candidate has fitness 128,500, giving it a roulette-wheel selection probability of about 0.39; the next strongest has fitness 90,000 and probability about 0.27. Neither is certain to be drawn: fitness changes the odds rather than fixing the result. section 7.3.3

Table 7.4 Selection probabilities (view table figure)

Candidate solutions x	Values of x in the solution space	Candidate solutions in the binary coding space	Objective function f(x)	Selection probability p_k
x₁	77	01001101	8,820	0.03
x₂	203	11001011	84,420	0.26
x₃	110	01101110	90,000	0.27
x₄	145	10010001	128,500	0.39
x₅	230	11100110	18,000	0.05

Compare each ticket-pricing candidate’s fitness with its fitness-proportionate selection probability. section 7.3.3

After reproduction, survivor selection fills the next population from existing individuals and new offspring. Options include random selection, removing the oldest individuals, fitness-proportionate replacement, elitist selection across old and new candidates, and tournaments. A generational genetic algorithm replaces the whole population with offspring; a steady-state algorithm generates fewer offspring and may replace only some existing individuals. Choose a method with both solution quality and population diversity in mind. section 7.3.5 section 7.3.4

How Optimization Algorithms: AI techniques for design, planning, and control problems covers it

Cover of Optimization Algorithms: AI techniques for design, planning, and control problems

liveBook manning.com

Read Optimization Algorithms: AI techniques for design, planning, and control problems on liveBook Get the book at manning.com

Also explained in

liveBook manning.com

Evolutionary Deep Learning: Genetic algorithms and neural networksA selection operator chooses individuals from a population for reproduction or survival according to their fitness.
liveBook manning.com

Advanced Algorithms and Data StructuresA selection technique is a genetic-algorithm method for choosing individuals that survive or reproduce in the next generation.
liveBook manning.com

Grokking AI Algorithms, Second EditionSelection is an evolutionary algorithm operation that chooses candidate solutions for continuation or reproduction based on their quality.
liveBook manning.com

Inside AISelection is a genetic-algorithm operation that retains candidate solutions with comparatively high fitness for further evolution.

Compare how 5 books explain selection →

Related concepts

measurementequivalent to · Measurement is the systematic assessment of a property, behavior, or outcome that produces an observed value, score, or judgment.
tournament selectionincludes · Tournament selection is an evolutionary technique that randomly samples groups of candidate solutions and selects the highest-fitness candidate from each group.
elitismincludes · Elitism is an evolutionary strategy that preserves high-performing individuals unchanged when forming the next generation.
fitnessuses · Fitness is a numerical score indicating how well a candidate solution performs relative to an optimization objective.
roulette-wheel selectionincludes · Roulette-wheel selection is an evolutionary selection technique that assigns candidates probabilities proportional to their fitness, typically giving lower-fitness candidates nonzero probabilities.
genetic algorithmused by · A genetic algorithm is an evolutionary optimization method that improves a population of candidate solutions through fitness-based selection and variation, often including recombination and mutation.

This explainer was generated by Manning's knowledge graph from the text of Optimization Algorithms: AI techniques for design, planning, and control problems. The section links point to the parts of the book it draws on.