3. CONTENTS
S.NO TOPICS
1 WHAT IS GENETIC ALGORITHM ?
2 HISTORY OF GENETIC ALGORITHM
3 WHAT IS THE USE OF GENETIC ALGORITHM ?
• HOW EFFECTIVE ARE GENETIC ALOGORITHM ?
4 HOW GENETICAL ALGORITHM WORKS
5 WHAT ARE THE TYPES OF GENETIC ALGORITHM ?
• Generational GA
• STEADY-STATE GA
• STEADY-GENERATIONAL GA
• (µ + µ)-GA
6 ADVANTAGES AND DISADVANTAGES OF GENTIC ALGORITHM
4. WHAT IS GENETIC ALGORITHM ?
The genetic algorithm is a method for solving both
constrained and unconstrained optimization problems that
is based on natural selection, the process that drives
biological evolution. The genetic algorithm repeatedly
modifies a population of individual solutions.
5. HISTORY OF GENETIC ALGORITHM
•As early as 1962, John Holland's work on adaptive systems laid the
foundation for later developments.
•By the 1975, the publication of the book Adaptation in Natural and
Artificial Systems, by Holland and his students and colleagues.
6. HISTORY OF GENETIC ALGORITHM
•In early to mid-1980s, genetic algorithms were being applied to a
broad range of subjects.
•In 1992 John Koza has used genetic algorithm to evolve programs
to perform certain tasks.
•He called his method "genetic programming" (GP).
7. WHAT IS THE USE OF GENETIC
ALGORITHM ?
Genetic algorithms are commonly used to generate high-
quality solutions to optimization and search problems by
relying on biologically inspired operators such as
mutation, crossover and selection.
8. • HOW EFFECTIVE ARE GENETIC
ALOGORITHM ?
Genetic Algorithms are good at taking large, potentially
huge search spaces and navigating them, looking for
optimal combinations of things, solutions you might not
otherwise find in a lifetime.
9. HOW GENETICAL ALGORITHM WORKS ?
The algorithm begins by creating a random initial population. The
algorithm then creates a sequence of new populations. At each
step, the algorithm uses the individuals in the current generation to
create the next population.
11. WHAT ARE THE TYPES OF GENETIC
ALGORITHM?
Four types of Genetic Algorithms (GA) are presented -
1]Generational GA (GGA), 2]Steady-State (µ + 1)-GA (SSGA),
3]Steady-Generational (µ, µ)-GA (SGGA), and 4](µ + µ)-GA. Based
on 30 runs of the best performing EC variants (a total of 12), each
crossover method for each type of GA is divided into its equivalent
classes.
12. 1.Generational GA
In the GGA, if the population size is P, there are P offspring that are
created and mutated. Following this, the replacement strategy
replaces all the parents with their offspring. This results in no
overlap between the current and new population. In this case,
elitism is 0. Here, the generation gap, which is the measure of
degree of overlap between the current and new generation, is zero.
13. 1.Generational GA
The GGA uses the tournament selection method to select the two
parents from the population to create one offspring. For each
offspring, two parents are randomly selected from the existing
population, and the process repeats until the number of offspring
reaches the current population size. Every member of the current
population is eliminated, and zero survivors remain.
14. 2.STEADY-STATE GA
The SSGA works by randomly selecting two parents, creating one
offspring, and replacing the worst fit individual in the population
with the offspring. The benefit of using a SSGA, rather than a GGA,
is that the SSGA makes only one function evaluation per child on
each cycle. A GGA must make P (where P is the population size)
function evaluations on each cycle.
15. 3.STEADY-GENERATIONAL GA
The SGGA works by first selecting the two parents and then
generates the offspring. Instead of the offspring replacing the
parents or the worst-fit individual, the offspring replaces a random
individual from the population that is not the best-fit. Similar to
the SSGA, the SGGA has the benefit of only making two function
evaluations on each cycle. The GGA requires P function evaluations
on each cycle.
16. 4. (µ + µ)-GA
The (µ + µ)-GA works by randomly selecting two parents with
binary tournament selection, creating an offspring, and adding the
offspring to a child population until the child population size is
equal to the original population size.
17. 4. (µ + µ)-GA
The arXiv:1911.00490v1 [cs.NE] 1 Nov 2019 algorithm then creates
a new population containing the original population and the child
population, and chooses the top individuals from this new
population until the population size is the same size as the original
population. In this way, the algorithm composes a population of
the most fit individuals out of two generations of individuals.
18. 4. (µ + µ)-GA
The benefit of using the (µ + µ)-GGA is that although there are
more function evaluations each cycle, the best fit individuals are
guaranteed in the new population as opposed to randomly
replacing individuals and potentially ending up with a a population
with lower fitnesses.
19. ADVANTAGES OF GENTIC ALGORITHM
1. The concept is easy to understand.
2. GA search from a population of points, not a single point.
3. GA use payoff (objective function) information, not derivatives.
4. GA supports multi-objective optimization.
20. ADVANTAGES OF GENTIC ALGORITHM
5. GA use probabilistic transition rules, not deterministic rules.
6. GA is good for "noisy" environments.
7. GA is robust w.r.t. to local minima/maxima.
8. GA is easily parallelised.
21. DISADVANTAGES OF GENETIC
ALGORITHM
1. GA implementation is still an art.
2. GA requires less information about the problem, but designing
an objective function and getting the representation and
operators right can be difficult.
3. GA is computationally expensive i.e. time-consuming.
