DREAM

DREAM
IDEA
IDEA
PLAN
PLAN
IMPLEMENTATION
IMPLEMENTATION
 Genetic algorithms
Present to:
Amirkabir University of Technology (Tehran
Polytechnic) & Semnan University

Dr. Kourosh Kiani

Email: kkiani2004@yahoo.com
Email: Kourosh.kiani@aut.ac.ir
Web: www.kouroshkiani.com
 Charles Darwin 1809 - 1882

"A man who dares to waste an hour of life has not discovered
the value of life"
 Genetic algorithms

Life on earth has evolved to be as it is through the

processes of natural selection, recombination and
mutation.
Genetic algorithms are a part of evolutionary computing,
which is a rapidly growing area of artificial intelligence.

As you can guess, genetic algorithms are inspired by Darwin's

theory of evolution. Simply said, problems are solved by an
evolutionary process resulting in a best (fittest) solution (survivor)
- in other words, the solution is evolved.
 History

Evolutionary computing was introduced in the 1960s by I.

Rechenberg in his work "Evolution strategies" (Evolutions
strategy in original). His idea was then developed by other
researchers. Genetic Algorithms (GAs) were invented by John
Holland and developed by him and his students and colleagues.
This lead to Holland's book "Adaption in Natural and Artificial
Systems" published in 1975.

In 1992 John Koza has used genetic algorithm to evolve

programs to perform certain tasks. He called his method
"genetic programming" (GP). LISP programs were used,
because programs in this language can expressed in the form of
a "parse tree", which is the object the GA works on.
 Biological Background : The cell
• Every animal cell is a complex of many small
“factories” working together
• The center of this all is the cell nucleus
• The nucleus contains the genetic information
 Biological Background : Chromosomes
• Genetic information is stored in the
chromosomes
• Each chromosome is build of DNA
• Chromosomes in humans form pairs
• There are 23 pairs
• The chromosome is divided in parts: genes
• Genes code for properties
• The possibilities of the genes for
one property is called: allele
• Every gene has an unique position
on the chromosome: locus
 Genetic algorithms
Every organism has a set of rules, describing how that organism is
built up from the tiny building blocks of life.
These rules are encoded in the genes of an organism, which in
turn are connected together into long strings called
chromosomes.
Each gene represents a specific characteristic of the organism,
like eye color or hair color, and has several different settings. For
example, the settings for a hair color gene may be blonde or
black. These genes and their settings are usually referred to as an
organism's genotype. The physical expression of the genotype -
the organism itself - is called the phenotype.
When two organisms mate they share their genes. The resultant
offspring may end up having half the genes from one parent and
half from the other. This process is called recombination. Very
occasionally a gene may be mutated. Normally this mutated gene
will not affect the development of the phenotype but very
occasionally it will be expressed in the organism as a completely
new characteristic .
 What Are Genetic
Algorithms?

►A way to employ evolution in the

computer
► Searchand optimization technique
based on variation and selection
Components of Genetic algorithms : Encoding

Encoding of chromosomes is the first question to ask when starting

to solve a problem with GA. Encoding depends on the problem
heavily.

This could be as a string of real numbers or, as is more typically the

case, a binary bit string.
Binary Encoding
Binary encoding is the most common one, mainly because the first
research of GA used this type of encoding and because of its
relative simplicity.
In binary encoding, every chromosome is a string of bits 0 or 1.
Chromosome 1011001011001010111001
A 01
Chromosome 1111111000001100000111
B 11
Components of Genetic algorithms : Encoding
Solutions are coded as “chromosomes”

Chromosome A 1 0 1 1 0 1 0 0 0 1 0 1 1

Size Shape Speed

Chromosome B 1 1 1 0 0 1 1 0 0 0 1 10 1

Size Shape Speed

Chromosome A 0 0 1 1 1 1 0 0 1 1 0 1 0 0 1

x y z

Chromosome B 1 1 1 1 0 1 0 1 1 1 0 1 0 0 0

x y z
Components of Genetic algorithms : Encoding

Given the digits 0 through 9 and the operators +, -, * and /,  find a
sequence that will represent a given target number. The operators will be
applied sequentially from left to right as you read.

So, given the target number 23, the sequence 6+5*4/2+1 would be one
possible solution.
If  75.5 is the chosen number then 5/2+9*7-5 would be a possible solution.

First we need to encode a possible solution as a string of bits… a

chromosome.
So how do we do this? Well, first we need to represent all the different
characters available to the solution... that is 0 through 9 and +, -, * and /.
This will represent a gene. Each chromosome will be made up of several
genes.
Components of Genetic algorithms : Encoding
Four bits are required to represent the range of characters used:
 
0: 0000
1: 0001
2: 0010 The possible genes 1110 & 1111 will remain
3: 0011 unused and will be ignored by the algorithm if
4: 0100 encountered.
5: 0101 So now you can see that the solution mentioned above
6: 0110 for 23, ' 6+5*4/2+1' would be represented by nine
7: 0111 genes like so:
8: 1000  
9: 1001 0110 1010 0101 1100 0100 1101 0010 1010 0001
+: 1010 6      +     5        *      4      /      2     +      1
-: 1011
*: 1100 These genes are all strung together to form the
/: 1101 chromosome:

  011010100101110001001101001010100001
Components of Genetic algorithms : Decoding
A Quick Word about Decoding
 
Because the algorithm deals with random arrangements of bits it is
often going to come across a string of bits like this:
 
0010001010101110101101110010
 
Decoded, these bits represent:
 
0010 0010 1010 1110 1011 0111 0010
2        2        +      n/a    -        7        2
 
Which is meaningless in the context of this problem! Therefore, when
decoding, the algorithm will just ignore any genes which don’t conform
to the expected pattern of: number -> operator -> number -> operator
…and so on. With this in mind the above ‘nonsense’ chromosome is
read (and tested) as:

0010 0010 1010 1110 1011 0111 0010

2        2        +     n/a    -        7        2
 
2   +   7
 Components of Genetic
algorithms : Decoding
0: 0000
1: 0001 0010 1010 0011 1100 0101 1010
2: 0010 0110
3: 0011 2        +        3     *    5      +      
4: 0100 6 =30
5: 0101
6: 0110
7: 0111 1000 1100 0011 1101 0011 1010
8: 1000 0010
9: 1001 8        *        3     /    3      +      
+: 1010 2 =10
-: 1011
*: 1100 0111 1010 0100 1011 0001 1101
/: 1101 0010
?        ?        ?     ?    ?       ?      
? =?
 Components of Genetic
algorithms :
Permutation encodingPermutation Encoding
can be used in ordering problems, such as travelling salesman
problem or task ordering problem.
In permutation encoding, every chromosome is a string of numbers that represent
a position in a sequence.
Chromosome A 1 5 3 2 6 4 7 9 8
Chromosome B 8 5 6 7 2 3 1 4 9

Example of chromosomes with permutation encoding

Permutation encoding is useful for ordering problems. For some types of crossover
and mutation corrections must be made to leave the chromosome consistent (i.e.
have real sequence in it) for some problems.

Example of Problem: Travelling salesman problem (TSP)

The problem: There are cities and given distances between them. Travelling
salesman has to visit all of them, but he does not want to travel more than necessary.
Find a sequence of cities with a minimal travelled distance.
Encoding: Chromosome describes the order of cities, in which the salesman will visit
them
 Components of Genetic
algorithms :
Value
Direct value encoding can be used Encoding
in problems where some more complicated values such as
real numbers are used. Use of binary encoding for this type of problems would be difficult.
In the value encoding, every chromosome is a sequence of some values. Values can be
anything connected to the problem, such as (real) numbers, chars or any objects.

Chromosome A 1.2324 5.3243 0.4556 2.3293 2.4545

Chromosome B ABDJEIFJDHDIERJFDLDFLFEGT
Chromosome C (back), (back), (right), (forward), (left)

Example of chromosomes with value encoding

Value encoding is a good choice for some special problems. However, for this encoding
it is often necessary to develop some new crossover and mutation specific for the
problem.

Example of Problem: Finding weights for a neural network

The problem: A neural network is given with defined architecture. Find weights between
neurons in the neural network to get the desired output from the network.
Encoding: Real values in chromosomes represent weights in the neural network.
 Crossover and Mutation
► Crossover: Operate on selected
parent chromosomes to create new
offspring.

► Mutation: Randomly changes the

offspring resulted from crossover to
prevent falling of all solutions into a
local optimum.
 Crossover

Crossover: single point - random

101000000 Parent1 Offspring1 1011011111

1011011111 Parent2 Offspring2 101000000
 Crossover
Single point cross-over

Multiple point cross-over

 Crossover

Crossover: two-point - random

parent 11010 offspring 1 1100 0

1
parent 10001 offspring 2 10 011
2
 Crossover
► Assign 'heads' to one parent, 'tails' to the other
► Flip a coin for each gene of the first child
► Make an inverse copy of the gene for the second
child
► Inheritance is independent of position

Crossover: Uniform
 Parent 1: 000000000000000000
 Parent 2: 111111111111111111
 Offspring 1: 010010110001011001
 Offspring 2: 101101001110100110
 Crossover
Single arithmetic
crossover
• Parents: 〈x1,…,xn 〉 and 〈y1,…,yn〉
• Pick a single gene (k) at random,
• child1 is: x1 , ..., xk , α ⋅ yk + (1 −α) ⋅ xk , ..., xn

• reverse for other child. e.g. with α = 0.8

Parent 1: 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Parent 2: 0.3 0.1 0.4 0.5 0.3 0.2 0.6 0.6 0.3

0.8 × 0.6 + (1 − 0.8)0.8 = 0.48 + 0.16 = 0.64 ≈ 0.6

Offspring 1: 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.9

0.8 × 0.8 + (1 − 0.8)0.6 = 0.64 + 0.12 = 0.76 ≈ 0.8

Offspring 1: 0.3 0.1 0.4 0.5 0.3 0.2 0.6 0.8 0.3
 Crossover
Simple arithmetic
Crossover
crossover
• Parents: 〈x1,…,xn 〉 and 〈y1,…,yn〉
• Pick random gene (k) after this point mix values
child1 is: x , ..., x , α ⋅ y
1 k k +1
+ (1 − α ) ⋅ x
k +1
, ..., α ⋅ y + (1 − α ) ⋅ x
n n

• reverse for other child. e.g. with α = 0.8

Parent 1: 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9

Parent 2: 0.3 0.1 0.4 0.5 0.3 0.2 0.6 0.6 0.3
0.8 × 0.6 + (1 − 0.8)0.8 = 0.48 + 0.16 = 0.64 ≈ 0.6
0.8 × 0.3 + (1 − 0.8)0.9 = 0.24 + 0.18 = 0.42 ≈ 0.4

Offspring 1: 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.6 0.4
0.8 × 0.8 + (1 − 0.8)0.6 = 0.64 + 0.12 = 0.76 ≈ 0.8
0.8 × 0.9 + (1 − 0.8)0.3 = 0.72 + 0.06 = 0.78 ≈ 0.8
Offspring 1: 0.3 0.1 0.4 0.5 0.3 0.2 0.6 0.8 0.8
 Crossover

None - no crossover, offspring is exact

copy of parents
 Mutation
mutated

Offspring1 1011011111 Offspring1 1011001111

Offspring2 1010000000 Offspring2 1000000000

Original offspring Mutated offspring

With some small probability (the mutation rate) flip each bit in the
offspring (typical values between 0.1 and 0.001)
 Mutation

Order changing mutation - two numbers are selected and

exchanged
(1 2 3 4 5 6 8 9 7) => (1 8 3 4 5 6 2 9 7)

Adding Mutation: a small number (for real value encoding) is

added to (subtracted from) selected values

(1.29  5.68  2.86  4.11  5.55) => (1.29  5.68  2.73  4.22  5.55)
Components of Genetic algorithms :
Create initial population (usually
random)
1 101100101100101011100101010100110110010110010101
110
2 001000101100101011100101010111110110010110000101
110
3 001000101100111111100101110111000010010110000101
110
4 100000101100101011101111110100110110010000010101
110
5 111111111100101011100101011111110110010110010111
111
6 001000101100111111111111111000110010010110000101
110
7 100000101001100111110010101110111111010011011001
001
8 111111111110000000001111101111110100110111111101
000
9 000000000000000000000000000000001111111101010101
011
10 000000000110011100001010101010100000000000011111
111
 Simple example of Genetic optimization

The chosen problem is to minimize the value of a string of 1’s and 0’s as a
binary number. While this is a trivial problem, it will serve to illustrate
the operation of the algorithm without the obscuring effects of a
complicated objective function.
The genetic algorithm performs minimization in the following steps:

1. A random population of chromosomes are generated.

2. The fitness function is applied to each chromosome.
3. The selection operation is applied favoring the fittest chromosomes.
4. The crossover operation is applied on the selected pairs to generate a new
population.
5. Mutation is applied to some offspring
6. Repeat steps 2-5
 Components of Genetic algorithms :
Selection of parents ( reproduction)
In this step the current population is used to produce a second generation
of population that, on the average, have a higher objective function.

Selection replicates the most successful solutions found in a population at

a rate proportional to their relative quality.

J chromosome Binary Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj)

number chromosome Objective function Fraction of Total

1 011010 26 0.0385 0.142

2 001011 11 0.0909 0.334
3 110110 54 0.0185 0.068
4 010011 19 0.0526 0.194
5 100111 39 0.0256 0.094
6 010110 22 0.0455 0.168
Total 0.2717 1
 Components of Genetic algorithms :
Selection of parents ( reproduction)
J Binary Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj) Round (Fj*6)=
chromosome chromosome Objective function Fraction of Total Copy number
number

1 011010 26 0.0385 0.142 1

2 001011 11 0.0909 0.334 2
3 110110 54 0.0185 0.068 0
4 010011 19 0.0526 0.194 1
5 100111 39 0.0256 0.094 1
6 010110 22 0.0455 0.168 1
Total 0.2717 1

Selected
In the reproduction population
0.142, 14%
phase, each chromosome 011010
0.168, 17%

is copied to the second 001011

01011
0
01101
0

generation a number of 0.094, 9% 10011

001011
times that is proportional 1
00101
1
to its fractional objective 010011 01001
1 0.334, 34%

function Fj. 100111 0.194, 19%

11011
0
010110 0.068, 7% Area is Proportional
to fitness value
 Components of Genetic algorithms :
Crossover (Recombination)
Recombination decomposes two distinct solutions (chromosomes)
and then randomly mixes their parts to form novel solutions
(chromosomes) .
This process is intended to simulate the exchange of genetic material
that occurs during biological reproduction. Here, pairs in the breeding
population are mated randomly. For simplicity we will mate adjacent
pairs. For each pair, a random integer from 1-6 (1- length of
chromosome) is chosen. This indicates how many bits on the right end
of each chromosome should be exchanged between the two mated
Before crossover:
chromosomes.

sP1 = 011010 sP3 = 001011 sP5 = 100111

sP2 = 001011 sP4 = 010011 sP6 = 010110

After crossover:
sC1 = 011011 sC3 = 000011 sC5 = 100110
sC2 = 001010 sC4 = 011011 sC6 = 010111
 Components of Genetic algorithms :
Mutation
A surprisingly small role is played by mutation. Biologically, it randomly perturbs the
population’s characteristics, thereby preventing evolutionary dead ends.
Most mutations are damaging rather than beneficial; therefore, rates must be low to avoid
the destruction of species. In the simulation we will assume a mutation rate of on bit per
thousand, where a mutation acts to reverse the value of single bit. Since we have only 36 bits
in our population, the probability of a mutation is only 0.036; therefore, no bits are reversed
in the simulation. By mutation, algorithm explores parts of the space that crossover might
miss, and helps prevent premature convergence.

Before 1101101101011 100

After 1101101001011100
 Components of Genetic algorithms :
Crossover (Recombination)

After One
J chromosome Binary Generation
Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj)
number chromosome Objective function Fraction of Total

1 011011 27 0.037 0.064

2 001010 10 0.10 0.174
3 000011 3 0.333 0.578
4 011011 27 0.037 0.064
5 100110 38 0.0263 0.046
6 010111 23 0.043 0.075
Total 0.576 1
 Components of Genetic algorithms :
Selection of parents ( reproduction)
J Binary Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj) Round (Fj*6)=
chromosome chromosome Objective function Fraction of Total Copy number
number

1 011011 27 0.037 0.064 0

2 001010 10 0.10 0.174 1

3 000011 3 0.333 0.578 4

4 011011 27 0.037 0.064 0

5 100110 38 0.0263 0.046 0

6 010111 23 0.043 0.075 1

Total 0.576 1

Selected
population 0.075, 7% 0.064, 6%
01011
0.046, 5% 01101
001010 1
In the reproduction 0.064, 6% 10011
1
01101
1
0.174, 17%

phase, each chromosome 000011 1

00101
0
is copied to the second 000011
generation a number of 000011 00001
1
times that is proportional 000011
to its fractional objective 0.578, 59%
Area is Proportional
010111 to fitness value
function Fj.
 Components of Genetic algorithms :
Crossover (Recombination)

Before crossover:
sP1 = 001010 sP3 = 000011 sP5 = 000011

sP2 = 000011 sP4 = 000011 sP6 = 010111

After crossover:
sC1 = 001011 sC3 = 000011 sC5 = 000011
sC2 = 000010 sC4 = 000011 sC6 = 010111
 Components of Genetic algorithms :
Crossover (Recombination)

After Two
J chromosome Binary Generation
Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj)
number chromosome Objective function Fraction of Total

1 001011 11 0.091 0.0557

2 000010 2 0.5 0.306
3 000011 3 0.333 0.204
4 000011 3 0.333 0.204
5 000011 3 0.333 0.204
6 010111 23 0.043 0.026
Total 1.633 1
 Components of Genetic algorithms :
Selection of parents ( reproduction)
J Binary Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj) Round (Fj*6)=
chromosome chromosome Objective function Fraction of Total Copy number
number

1 001011 11 0.091 0.0557 0

2 000010 2 0.5 0.306 2

3 000011 3 0.333 0.204 1

4 000011 3 0.333 0.204 1

5 000011 3 0.333 0.204 1

6 010111 23 0.043 0.026 0

Total 1.633 1
Selected
population 0.026, 3% 0.0557, 6%

000010 0.204, 20%

In the reproduction 000010 0.306, 31%

phase, each chromosome
000011
is copied to the second
000011
generation a number of 0.204, 20%

times that is proportional 000011

0.204, 20%
to its fractional objective 000011 Area is Proportional
to fitness value
 Components of Genetic algorithms :
Crossover (Recombination)

Before crossover:
sP1 = 000010 sP3 = 000011 sP5 = 000011

sP2 = 000010 sP4 = 000011 sP6 = 000011

After crossover:
sC1 = 000010 sC3 = 000011 sC5 = 000011
sC2 = 000010 sC4 = 000011 sC6 = 000011
 Components of Genetic algorithms :
Crossover (Recombination)

After Three
J chromosome Binary Generation
Decimal value Oj = 1/(decimal value) Fj =Oj / (Total Oj)
number chromosome Objective function Fraction of Total

1 000010 2 0.5 0.214

2 000010 2 0.5 0.214
3 000011 3 0.333 0.143
4 000011 3 0.333 0.143
5 000011 3 0.333 0.143
6 000011 3 0.333 0.143
Total 16 2.332 1
 The Evolutionary Cycle

Create initial random population

Evaluate fitness of each individual

yes
Termination criteria satisfied ? stop
no
Select parents according to fitness

Recombine parents to generate offspring

Mutate offspring

Replace population by new offspring

 Example: the MAXONE problem
Suppose we want to maximize the number of ones
in a string of l binary digits
Is it a trivial problem?
It may seem so because we know the answer in advance

► An individual is encoded (naturally) as a string of l binary

digits
► The fitness f of a candidate solution to the MAXONE problem
is the number of ones in its genetic code
► We start with a population of n random strings. Suppose that
l = 10 and n = 6
 Genetic algorithms : Example
(initialization)
the MAXONE problem

J chromosome Binary Oj = number of ones Fj =Oj / (Total Oj)

number chromosome Objective function Fraction of Total

1 1111010101 7 0.206

2 0111000101 5 0.147

3 1110110101 7 0.206

4 0100010011 4 0.118

5 1110111101 8 0.235

6 0100110000 3 0.088

Total 34 1
 Genetic algorithms :
Selection of parents ( reproduction)
J Binary Oj = 1/(decimal value) Fj =Oj / (Total Oj) Round (Fj*6)=
chromosome chromosome Objective function Fraction of Total Copy number
number

1 1111010101 7 0.206 1
2 0111000101 5 0.147 1
3 1110110101 7 0.206 1
4 0100010011 4 0.118 1
5 1110111101 8 0.235 1
6 0100110000 3 0.088 0
Total 34 1

Selected
population

1111010101

0111000101

1110110101

0100010011
 Genetic algorithms : Example
(crossover1)
the MAXONE problem

Next we mate strings for crossover. For each couple

we decide according to crossover probability (for
instance 0.6) whether to actually perform crossover
or not.
Suppose that we decide to actually perform
crossover only for couples (s1`, s2`) and (s5`, s6`). For
each couple, we randomly extract a crossover point,
for instance 2 for the first and 5 for the second
Genetic algorithms : Example (crossover
2)
the MAXONE problem
Before crossover:
sP1 = 1111010101 sp4 = 0100010011
sP3 = 1110110101 sP5 = 1110111101

After crossover:
sC1 = 1110110101 sC4 = 0100011101
sC3 = 1111010101 sC5 = 1110110011
 Genetic algorithms : Example (Mutation)

the MAXONE problem

The final step is to apply random mutation: for each bit that we are to
copy to the new population we allow a small probability of error (for
instance 0.1)

Before applying mutation: After applying mutation:

Modified Modified
population population

1110110101 1110100101

0111000101 0111000101

1111010101 1111110100

0100011101 0100011101

1110110011 1110110001
 Genetic algorithms : Example (End)
the MAXONE problem
J chromosome Binary Oj = number of ones Fj =Oj / (Total Oj)
number chromosome Objective function Fraction of Total

1 1110100101 6 0.162

2 0111000101 5 0.135

3 1111110100 7 0.189

4 0100011101 5 0.135

5 1110110001 6 0.162
In one generation, the total population fitness changed from 34
to 37, thus
6 improved by ~9%
1110101111 8 0.216

At this point,
Total we go through the same
37 process all 1over again,
until a stopping criterion is met
 Genetic algorithms : Optimization of a
simple function
The function is defined as:

f ( x) = x. sin(10π .x) + 1.0

3

2.5

The problem is to find x from the rang [-1 2] 1.5

which 1

Maximizes the function f, i.e., to find x0 such

f ( x ) ≥ f ( x), for all x ∈ [−1 2] 0.5

that,0 0

It is relatively easy to analyze the function f. -0.5

The zeros of the first derivative f’ should be -1

-1 -0.5 0 0.5 1 1.5 2

determined:
f ( x) = sin(10π .x) + 10πx. cos(10π .x) = 0 The function f reaches its local
tan(10π .x) = −10π .x max for xi if i is an odd integer and
2i − 1 its local min for xi if i is an even
xi = for i = 1,2,  integer. The function reaches its
20
max for:
37
x0 = 0 x19 = = 1.85
20
2i + 1 π
xi = for i = −1,−2, f (1.85) = 1.85. sin(18π + ) + 1.0 = 2.85
20 2
 Genetic algorithms : Optimization of a
simple function

Assume that we wish to construct a

genetic algorithm: to maximize the
function f.
Let discuss the major components
of such a genetic algorithm in turn.
 Genetic algorithms : Presentation
(Encoding)
We use a binary vector as a chromosome to present real values of the
variable x.
The length of the vector depends on the required precision, which, in this
example, is six places after the decimal point.
The domain of the variable x has length 3; the precision requirement
implies that the range [-1 2] should be divided into at least
3.1,000,000=3,000,000 equal size ranges. This means that 22 bits are
required as a binary2097152 = 2 < 3000000 < 2 = 4194304
21
vector (chromosome):
22

he mapping from a binary vector (chromosome) (b21 b20 ....b0) into a real number
from the range [-1 2] is straightforward and is completed in to steps:
Convert the binary chromosome (b21 b20 ….b0) from the base 2 to base 10:
 21 
(b21b20  b0 ) 2 =  ∑ bi .2i  = x '
 i =0 10
• Find a corresponding real number x :

bi − ai 3
xi = ai + decimal(10001....100002 ). x = −1.0 + x ' .
2 mi − 1 2 22 − 1

Where -1.0 is the left boundary of the domain and 3 is the length of the domain:
 Genetic algorithms : Presentation
(Encoding)
Chromosome: 1000101110110101000111

x ' = (1000101110110101000111) 2 = 2288967

3
x = −1.0 + x . 22
'

2 −1
3
x = −1.0 + 2288967. = 0.637197
4194303

Of course, the chromosomes (0000000000000000000000) and (111111111111111111111

Represent boundaries of the domain, -1.0 and 2.0
 Genetic algorithms : Initial
population

The initialization process is very simple:

we create a population of chromosomes,
where each chromosome is is a binary
vector of 22 bits. All 22 bits for each
chromosome are initialized
Randomly.
 Genetic algorithms :
Evaluation function
Evaluation function eval(*) for binary chromosome v is equivalent to the function f:

eval (v) = f ( x)
As noted earlier, the evaluation plays the role of the environment, rating potential solutio
In terms of their fitness. For example, three chromosomes:
v1 = (1000101110110101000111) corresponed to x1 = 0.637197
v2 = (0000001110000000010000) corresponed to x2 = −0.958973
v3 = (1110000000111111000101) corresponed to x3 = 1.627888

eval (v1 ) = f ( x1 ) = 1.586345

eval (v2 ) = f ( x2 ) = 0.078878
eval (v3 ) = f ( x3 ) = 2.250650

learly, the chromosome v3 is the best of the three chromosomes, since its evaluation retur
he highest value.
 Genetic algorithms :
Mutation
Assume that the fifth gene from the v3 chromosome was selected for mutation:

v3 = (1110000000111111000101) corresponed to x3 = 1.627888

v '3 = (1110100000111111000101) corresponed to x '3 = 1.721638

eval (v3 ) = f ( x3 ) = 2.250650

eval (v '3 ) = f ( x '3 ) = −0.082257
This means that this particular mutation resulted in a significant decrease of the value
the chromosome v3. If the 10th gene was selected for mutation in chromosome v3, then
v3 = (1110000000111111000101) corresponed to x3 = 1.627888
v"3 = (1110000001111111000101) corresponed to x"3 = 1.630818
eval (v3 ) = f ( x3 ) = 2.250650
eval (v "3 ) = f ( x"3 ) = 2.343555
 Genetic algorithms :
Crossover
Before crossover:
v2 = 0000001110000000010000
v3 = 1110000000111111000101
v2 = (0000001110000000010000) corresponed to x2 = −0.958973 eval (v2 ) = f ( x2 ) = 0.078878
v3 = (1110000000111111000101) corresponed to x3 = 1.627888 eval (v3 ) = f ( x3 ) = 2.250650

After crossover:
v’2 = 0000000000111111000101
v’3 = 1110001110000000010000
v ' 2 = (0000000000111111000101) corresponed to x ' 2 = −0.998113 eval (v ' 2 ) = f ( x ' 2 ) = 0.940865
v '3 = (1110001110000000010000) corresponed to x '3 = 1.666028 eval (v '3 ) = f ( x '3 ) = 2.459245

Note that the second offspring has a better evaluation than both of its parents.
 Genetic algorithms :
Parameters
For this particular problem we have used the following parameters: Population
size: pop_size = 50, probability of crossover Pc =0.25, probability of mutation Pm
= 0.01.
Generation Evaluation function
number
In table, we provide the generation
number for which we noted an 1 1.441942
improvement in the evaluation function, 6 2.250003
together with the value of function. The
8 2.250283
best chromosome after 150 generations
was: 9 2.250284
10 2.250363
vmax = (1111001101000100000101) 12 2.328077
corresponed to xmax = 1.850773 39 2.344251
40 2.345087
51 2.738930
99 2.849246
137 2.850217
145 2.850227
 Genetic algorithms : Optimization of a two
variables function
The function is defined as: f ( x1 , x2 ) = 21.5 + x1. sin( 4π .x1 ) + + x2 . sin( 20π .x2 )

Where − 3 ≤ x1 ≤ 12.1 and 4.1 ≤ x2 ≤ 5.8

 Genetic algorithms : Presentation
(Encoding)
We use a binary vector as a chromosome to present real values of the
variable x.
The length of the vector depends on the required precision, which, in this
example, is four places after the decimal point.

The domain of the variable x1 has length 15.1; the precision requirement
implies that the range [-3 12.1] should be divided into at least
15.1*10000=151000 equal size ranges. This means that 18 bits are
required as a binary vector (chromosome):
131072 = 2 < 151000 < 2 = 262144
17 18

The domain of the variable x2 has length 1.7; the precision requirement
implies that the range [4.1 5.8] should be divided into at least
1.7*10000=17000 equal size ranges. This means that 15 bits are required
as a binary vector (chromosome):

16384 = 214 < 17000 < 215 = 32768

 Genetic algorithms : Presentation
(Encoding)
The total length of a chromosome (solution vector) is then m = 18 +
15=33 bits; the first 18 bits code x1 and remaining 15 bits (19-33) code x2.
Let us consider an example chromosome:
010001001011010000111110010100010
bi − ai
xi = ai + decimal (10001....100002 ).
2 mi − 1

The first 18 bits: 010001001011010000 represents

12.1 − (−3.0) 15.1
x1 = −3.0 + decimal (0100010010 11010000 2 ). = −3. 0 + 70352 . = 1.052426
2 −1
18
262143

The first 18 bits: 111110010100010 represents:

5.8 − 4.1 1.7
x2 = 4.1 + decimal (1111001010 0010 2 ). = 4.1 + 31906 . = 5.755330
2 −1
15
32767
So the chromosome: 010001001011010000111110010100010
corresponds to (x1, x2)= (1.052426,5.755330). The fitness value for this
chromosome is:

f(1.052426,5.755330)=20.252640.
 Genetic algorithms : Initial
population
To optimize the function f using a genetic algorithm, we create a
population of pop_size = 20 chromosomes. All 33 bits in all chromosomes
are initialized randomly.
v1 = 100110100000001111111010011011111
v2 = 111000100100110111001010100011010
v3 = 000010000011001000001010111011101
v4 = 100011000101101001111000001110010
v5 = 000111011001010011010111111000101
v6 = 000101000010010101001010111111011
v7 = 001000100000110101111011011111011
v8 = 100001100001110100010110101100111
v9 = 010000000101100010110000001111100
v10 = 000001111000110000011010000111011
v11 = 011001111110110101100001101111000
v12 = 110100010111101101000101010000000
v13 = 111011111010001000110000001000110
v14 = 010010011000001010100111100101001
v15 = 111011101101110000100011111011110
v16 = 110011110000011111100001101001011
v17 = 011010111111001111010001101111101
v18 = 011101000000001110100111110101101
v19 = 000101010011111111110000110001100
v20 = 101110010110011110011000101111110
During the evaluation phase we decode each chromosome and calculate
the fitness
Function values from (x1, x2) values just decoded. We get:
v1 = 100110100000001111111010011011111 | x1 = 6.0844 x2 = 5.6522 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 26.0196
v2 = 111000100100110111001010100011010 | x1 = 10.3484 x2 = 4.3802 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 7.5800
v3 = 000010000011001000001010111011101 | x1 = −2.5166 x2 = 4.3903 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 19.5263
v4 = 100011000101101001111000001110010 | x1 = 5.2786 x2 = 5.5934 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 17.4067
v5 = 000111011001010011010111111000101 | x1 = −1.2551 x2 = 4.7344 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 25.3411
v6 = 000101000010010101001010111111011 | x1 = −1.8117 x2 = 4.3919 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 18.1004
v7 = 001000100000110101111011011111011 | x1 = −0.9914 x2 = 5.5802 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 16.0208
v8 = 100001100001110100010110101100111 | x1 = 4.9106 x2 = 4.7030 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 17.9597
v9 = 010000000101100010110000001111100 | x1 = 0.7954 x2 = 5.3814 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 16.1277
v10 = 000001111000110000011010000111011 | x1 = −2.5548 x2 = 4.7937 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 21.2784
v11 = 011001111110110101100001101111000 | x1 = 3.1300 x2 = 4.9960 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 23.4106
v12 = 110100010111101101000101010000000 | x1 = 9.3561 x2 = 4.2394 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 15.0116
v13 = 111011111010001000110000001000110 | x1 = 11.1346 x2 = 5.3786 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 27.3167
v14 = 010010011000001010100111100101001 | x1 = 1.3359 x2 = 5.1513 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 19.8762
v15 = 111011101101110000100011111011110 | x1 = 11.0889 x2 = 5.0545 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 30.0602
v16 = 110011110000011111100001101001011 | x1 = 9.2115 x2 = 4.9937 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 23.8672
v17 = 011010111111001111010001101111101 | x1 = 3.3675 x2 = 4.5713 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 13.6961
v18 = 011101000000001110100111110101101 | x1 = 3.8430 x2 = 5.1582 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 15.4141
v19 = 000101010011111111110000110001100 | x1 = −1.7466 x2 = 5.3955 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 20.0959
v20 = 101110010110011110011000101111110 | x1 = 7.9359 x2 = 4.7573 | eval ( x1 , x2 ) = f ( x1 , x2 ) = 13.6669
Total Fitness 387.7768

It is clear, that the chromosome v15 is the strongest one, and the
chromosome v2 the weakest.
Now we constructs a roulette wheel for the selection process. The total
20
fitness of the population
F = evalis:
∑ (v ) = 387.7768
i =1
i

The probability of a selection pi and the cumulative probabilities qi for

each chromosome vi (i=1,…20) are:
p1 = eval (v1 ) / F = 26.0196 / 387.6822 = 0.0671 q1 = 0.0671
p2 = eval (v2 ) / F = 7.5800 / 387.6822 = 0.0195 q2 = 0.0866
p3 = eval (v3 ) / F = 19.5263 / 387.6822 = 0.0504 q3 = 0.1370
p4 = eval (v4 ) / F = 17.4067 / 387.6822 = 0.0449 q4 = 0.1819
p5 = eval (v5 ) / F = 25.3411 / 387.6822 = 0.0654 q5 = 0.2472
p6 = eval (v6 ) / F = 18.1004 / 387.6822 = 0.0467 q6 = 0.2939
p7 = eval (v7 ) / F = 16.0208 / 387.6822 = 0.0413 q7 = 0.3352
p8 = eval (v8 ) / F = 17.9597 / 387.6822 = 0.0463 q8 = 0.3815
p9 = eval (v9 ) / F = 16.1277 / 387.6822 = 0.0416 q9 = 0.4231
p10 = eval (v10 ) / F = 21.2784 / 387.6822 = 0.0549 q10 = 0.4780
p11 = eval (v11 ) / F = 23.4106 / 387.6822 = 0.0604 q11 = 0.5384
p12 = eval (v12 ) / F = 15.0116 / 387.6822 = 0.0387 q12 = 0.5771
p13 = eval (v13 ) / F = 27.3167 / 387.6822 = 0.0775 q13 = 0.6475
p14 = eval (v14 ) / F = 19.8762 / 387.6822 = 0.0513 q14 = 0.6988
p15 = eval (v15 ) / F = 30.0602 / 387.6822 = 0.0775 q15 = 0.7763
p16 = eval (v16 ) / F = 23.8672 / 387.6822 = 0.0615 q16 = 0.8379
p17 = eval (v17 ) / F = 13.6961 / 387.6822 = 0.0353 q17 = 0.8732
p18 = eval (v18 ) / F = 15.4141 / 387.6822 = 0.0397 q18 = 0.9129
p19 = eval (v19 ) / F = 20.0959 / 387.6822 = 0.0518 q19 = 0.9648
p20 = eval (v20 ) / F = 13.6669 / 387.6822 = 0.0352 q20 = 1.0000
 Now we are ready to spin the q10 = 0.4780 < r1 = 0.5138 < q11 = 0.5384 Select v11 chromosome
roulette wheel 20 times; each q3 = 0.1370 < r2 = 0.1758 < q4 = 0.1819 Select v4 chromosome
time we select a single
q6 = 0.1717 < r3 = 0.3087 < q7 = 0.1819 Select v7 chromosome
chromosome for a new population.
Let us assume that a (random) q10 = 0.4780 < r4 = 0.5345 < q11 = 0.5384 Select v11 chromosome
sequence of 20 numbers from q18 = 0.9129 < r5 = 0.9476 < q19 = 0.9648 Select v19 chromosome
r1range
= 0.5138 [0 1]
q1 =is:
0.0670 v =11v '1 = v 11 q3 = 0.1370 < r6 = 0.1717 < q4 = 0.1819 Select v4 chromosome
r2 = 0.1758 q2 = 0.0866 v2 = 0 v 2 = v4
'

v3 = 0
q14 = 0.6988 < r7 = 0.7022 < q15 = 0.7763 Select v15 chromosome
r3 = 0.3087 q3 = 0.1370 v '3 = v7
r4 = 0.5345 q4 = 0.1819 v4 = 2 v ' 4 = v11 q4 = 0.1819 < r8 = 0.2226 < q5 = 0.2472 Select v5 chromosome
r5 = 0.9476 q5 = 0.2472 v5 = 1 v '5 = v19 q10 = 0.4780 < r9 = 0.4948 < q11 = 0.5384 Select v11 chromosome
r6 = 0.1717 q6 = 0.2939 v6 = 1 v 6 = v4
'

v7 = 1
q9 = 0.4231 < r10 = 0.4247 < q10 = 0.4780 Select v10 chromosome
r7 = 0.7022 q7 = 0.3352 v ' 7 = v15
r8 = 0.2226 q8 = 0.3815 v8 = 1 v '8 = v5 q14 = 0.6988 < r11 = 0.7039 < q15 = 0.7763 Select v15 chromosome
r9 = 0.4948 q9 = 0.4231 v9 = 1 v '9 = v11 q8 = 0.3815 < r12 = 0.3890 < q9 = 0.4231 Select v9 chromosome
r10 = 0.4247 q10 = 0.4780 v10 = 1 v '10 = v10
v11 = 3
q5 = 0.2472 < r13 = 0.2772 < q6 = 0.2939 Select v6 chromosome
r11 = 0.7039 q11 = 0.5384 v '11 = v15

r12 = 0.3890 q12 = 0.5771 v12 = 0 v '12 = v9 q7 = 0.3352 < r14 = 0.3681 < q8 = 0.3815 Select v8 chromosome
r13 = 0.2772 q13 = 0.6475 v13 = 1 v '13 = v6 q19 = 0.9648 < r15 = 0.9834 < q20 = 1.0000 Select v20 chromosome
v14 = 0 v '
= v8
r14 = 0.3681 q14 = 0.6988 14
q0 = 0.0000 < r16 = 0.0054 < q1 = 0.0670 Select v1 chromosome
v15 = 4 v '15 = v20
r15 = 0.9834 q15 = 0.7763
v16 = 1 v '16 = v1 q14 = 0.6988 < r17 = 0.7656 < q15 = 0.7763 Select v15 chromosome
r16 = 0.0054 q16 = 0.8379
v17 = 0 v '17 = v15 q12 = 0.5771 < r18 = 0.6446 < q13 = 0.6475 Select v13 chromosome
r17 = 0.7656 q17 = 0.8732
v18 = 0 v '
= v13
r18 = 0.6446 q18 = 0.9129 18
q14 = 0.6988 < r19 = 0.7671 < q15 = 0.7763 Select v15 chromosome
v19 = 1 v '19 = v15
r19 = 0.7671 q19 = 0.9648
v20 = 1 v ' 20 = v16 q15 = 0.7763 < r20 = 0.7802 < q16 = 0.8379 Select v16 chromosome
r20 = 0.7802 q20 = 1.0000
Genetic algorithms : Crossover
Now we are ready to apply the crossover operation to the individuals in the new
population (vector v’I ). The probability of crossover Pc=0.25, so we expect that on
average 25% of chromosomes (i.e., 4 out of 20) undergo crossover.
We proceed in the following way: Let us assume that a (random) sequence of 20
numbers from range [0 1] is:
r1 =0.8230
r2 =0.1519 This mean that the chromosomes v’2, v’11 , v’13 and v’18 were selected
r3 =0.6255
for crossover. For each of these two pairs, we generate a random
r4 =0.3115
r5 =0.3490
integer number pos from range [1 32] ( 33 is the total length-
r6 =0.9172 number of bits-in chromosome). The number pos indicates the
r7 =0.5198 position of the crossing point. The first pair of chromosomes is:
r8 =0.4012
r9 =0.6068
v ' 2 = 100011000101101001111000001110010
r10 =0.7854 v '11 = 111011101101110000100011111011110
r11 =0.0315
r12 =0.8699 And the generated number
r13 =0.1665 pos=9.
r14 =0.6745 v 2 = 100011000101110000100011111011110
''

r15 =0.7584
v ''11 = 111011101101101001111000001110010
r16 =0.5819
r17 =0.3892
r18 =0.2002
r19 =0.3557
r20 =0.8269
Genetic algorithms : Crossover
The second pair of chromosomes is: The current version of the
population
And the generated number v 1 = 0110011111 1011010110 0001101111 000
'

pos=9. v '' 2 = 1000110001 0111000010 0011111011 110

v '13 = 000101000010010101001010111111011 v '3 = 0010001000 0011010111 1011011111 011
v ' 4 = 0110011111 1011010110 0001101111 000
v '
18 = 111011111010001000110000001000110 v '5 = 0001010100 1111111111 0000110001 100
v ' 6 = 1000110001 0110100111 1000001110 010
v ' 7 = 1110111011 0111000010 0011111011 110
v ''13 = 000101000010010101000000001000110 v '8 = 0001110110 0101001101 0111111000 101
v '9 = 0110011111 1011010110 0001101111 000
v ''
18 = 111011111010001000111010111111011
v '10 = 0000011110 0011000001 1010000111 011
v ''11 = 1110111011 0110100111 1000001110 010
v '12 = 0100000001 0110001011 0000001111 100
v ''13 = 0001010000 1001010100 0000001000 110
v '14 = 1000011000 0111010001 0110101100 111
v '15 = 1011100101 1001111001 1000101111 110
v '16 = 1001101000 0000111111 1010011011 111
v '17 = 1110111011 0111000010 0011111011 110
v ''18 = 1110111110 1000100011 1010111111 011
v '19 = 1110111011 0111000010 0011111011 110
v ' 20 = 1100111100 0001111110 0001101001 011
Genetic algorithms : Mutation
The next operator, mutation, is performed on a bit-by-bit basis. The probability of
mutation pm=0.01, so we expect that (on average) 1% of bits would undergo
mutation. There are
m x pop_size =33 x 20 = 660 bits in the whole population; we expect (on average)
6.6 mutations per generation. Every bit has an equal chance to be mutated, so, for
every bit in the population, we generate a random number r from the range [0 1];
if r<0.01, we mutate the bit.
This means that we have to generate 660 random numbers. In a sample run, 5 of
these numbers were smaller than 0.01; the bit number and the random number
are listed below:
Bit position Random number Bit position Chromosome Bit number within
position chromosome
112 0.0002 112 int(112/33) = 4 112-3x33 = 13
349 0.0099 349 int(349/33) = 11 349-10x33 = 19
418 0.0088
418 int(418/33) = 13 418-12x33 = 22
429 0.0054
602 0.0028
429 int(429/33) = 13 429-12x33 = 33

602 int(602/33) = 19 602-18x33 = 8

 Genetic algorithms : Mutation
The current version of the The current version of the
population before mutation population after mutation
v '1 = 0110011111 1011010110 0001101111 000 v '1 = 0110011111 1011010110 0001101111 000

Chromosome position

Bit number within chromosome

v '' 2 = 1000110001 0111000010 0011111011 110 v '' 2 = 1000110001 0111000010 0011111011 110
v '3 = 0010001000 0011010111 1011011111 011 v '3 = 0010001000 0011010111 1011011111 011
v ' 4 = 0110011111 1011010110 0001101111 000 v ' 4 = 0110011111 1001010110 0001101111 000
v '5 = 0001010100 1111111111 0000110001 100 v '5 = 0001010100 1111111111 0000110001 100
v ' 6 = 1000110001 0110100111 1000001110 010 v ' 6 = 1000110001 0110100111 1000001110 010
v ' 7 = 1110111011 0111000010 0011111011 110 v ' 7 = 1110111011 0111000010 0011111011 110
v '8 = 0001110110 0101001101 0111111000 101 v '8 = 0001110110 0101001101 0111111000 101
v '9 = 0110011111 1011010110 0001101111 000 v '9 = 0110011111 1011010110 0001101111 000
v '10 = 0000011110 0011000001 1010000111 011 v '10 = 0000011110 0011000001 1010000111 011
v ''11 = 1110111011 0110100111 1000001110 010 v ''11 = 1110111011 0110100101 1000001110 010
v '12 = 0100000001 0110001011 0000001111 100 v '12 = 0100000001 0110001011 0000001111 100
v ''13 = 0001010000 1001010100 0000001000 110 4 13 v ''13 = 0001010000 1001010100 0100001000 111
v '14 = 1000011000 0111010001 0110101100 111 11 19 v '14 = 1000011000 0111010001 0110101100 111
v '15 = 1011100101 1001111001 1000101111 110 v '15 = 1011100101 1001111001 1000101111 110
13 22
v '16 = 1001101000 0000111111 1010011011 111 v '16 = 1001101000 0000111111 1010011011 111
13 33
v '17 = 1110111011 0111000010 0011111011 110 v '17 = 1110111011 0111000010 0011111011 110
v ''18 = 1110111110 1000100011 1010111111 011 19 8 v ''18 = 1110111110 1000100011 1010111111 011
v '19 = 1110111011 0111000010 0011111011 110 v '19 = 1110111111 0111000010 0011111011 110
v ' 20 = 1100111100 0001111110 0001101001 011 v ' 20 = 1100111100 0001111110 0001101001 011
We drop primes for modified chromosomes : the
population is listed as new vector vi:
eval (v1 ) = f (3.1301, 4.9961 ) = 23 .4107
v1 = 0110011111 1011010110 0001101111 000
eval (v2 ) = f (5.2790 , 5.0545 ) = 18 .2011
v2 = 1000110001 0111000010 0011111011 110
eval (v3 ) = f ( −0.9915 , 5.6803 ) = 16 .0208
v3 = 0010001000 0011010111 1011011111 011
eval (v4 ) = f (3.1282 , 4.9961 ) = 23 .4126
v4 = 0110011111 1001010110 0001101111 000
eval (v5 ) = f (−1.7466 , 5.3956 ) = 20 .0959
v5 = 0001010100 1111111111 0000110001 100
eval (v6 ) = f (5.2787 , 5.5935 ) = 17 .4067
v6 = 1000110001 0110100111 1000001110 010
eval (v7 ) = f (11 .0890 , 5.0545 ) = 30 .0602
v7 = 1110111011 0111000010 0011111011 110
eval (v8 ) = f ( −1.2552 , 4.7345 ) = 25 .3412
v8 = 0001110110 0101001101 0111111000 101
eval (v9 ) = f (3.1301, 4.9961 ) = 23 .4107
v9 = 0110011111 1011010110 0001101111 000
eval (v10 ) = f (−2.5549 , 4.7937 ) = 21 .2784
v10 = 0000011110 0011000001 1010000111 011
eval (v11 ) = f (11 .0886 , 4.7434 ) = 33 .3518
v11 = 1110111011 0110100101 1000001110 010
eval (v12 ) = f (0.7954 , 5.3815 ) = 16 .1278
v12 = 0100000001 0110001011 0000001111 100
eval (v13 ) = f (−1.8118 , 4.2099 ) = 22 .6925
v13 = 0001010000 1001010100 0100001000 111
eval (v14 ) = f (4.9106 , 4.7030 ) = 17 .9597
v14 = 1000011000 0111010001 0110101100 111
eval (v15 ) = f (7.9360 , 4.7573 ) = 13 .6670
v15 = 1011100101 1001111001 1000101111 110
eval (v16 ) = f (6.0844 , 5.6522 ) = 26 .0196
v16 = 1001101000 0000111111 1010011011 111
eval (v17 ) = f (11 .0890 , 5.0545 ) = 30 .0602
v17 = 1110111011 0111000010 0011111011 110
eval (v18 ) = f (11 .1346 , 5.6670 ) = 27 .5911
v18 = 1110111110 1000100011 1010111111 011
eval (v19 ) = f (11 .0595 , 5.0545 ) = 27 .6084
v19 = 1110111111 0111000010 0011111011 110
eval (v20 ) = f (9.2116 , 4.9937 ) = 23 .8672
v20 = 1100111100 0001111110 0001101001 011
Total = 447 .0497
 Now, we are ready to run the selection process again and apply
the genetic operator, evaluate the next generation, etc. After 100
generations the population is:

v1 = 1110111101 1001101110 0101010111 011 eval (v1 ) = f (11 .12, 5.0925 ) = 30 .2985
v2 = 1110011001 1000010001 0101010111 000 eval (v2 ) = f (10 .5888 , 4.6674 ) = 26 .8697
v3 = 1110111101 1101101110 0101010111 011 eval (v3 ) = f (11 .1246 , 5.0925 ) = 30 .3166
v4 = 1110011000 1000011000 0101010111 001 eval (v4 ) = f (10 .5741, 4.2442 ) = 31 .9331
v5 = 1110111101 1101101110 0101010111 011 eval (v5 ) = f (11 .1246 , 5.0925 ) = 30 .3166
v6 = 1110011001 1000010000 0100010100 001 eval (v6 ) = f (10 .5888 , 4.2146 ) = 34 .3561
v7 = 1101011000 1001001000 1100010110 000 eval (v7 ) = f (9.6311, 4.4279 ) = 35 .4586
v8 = 1111011000 1000101000 1101010010 001 eval (v8 ) = f (11 .5181, 4.4528 ) = 23 .3090
v9 = 1110011000 1001001000 1100010110 001 eval (v9 ) = f (10 .5748 , 4.4279 ) = 34 .3938
v10 = 1110111101 1101101110 0101010111 011 eval (v10 ) = f (11 .1225 , 5.0925 ) = 30 .3166
v11 = 1101011000 0001001000 1100010110 000 eval (v11 ) = f (9.6237 , 4.4279 ) = 35 .4779
v12 = 1101011000 1001001000 1100010110 001 eval (v12 ) = f (9.6311, 4.42779 ) = 35 .4561
v13 = 1110111101 1101101110 0101010111 011 eval (v13 ) = f (11 .1246 , 5.0925 ) = 30 .3166
v14 = 1110011001 1000010000 0101010111 011 eval (v14 ) = f (10 .5888 , 4.2425 ) = 32 .9321
v15 = 1110011010 1011100101 0100110110 001 eval (v15 ) = f (10 .6066 , 4.6537 ) = 30 .7468
v16 = 1110011001 1000010100 0100010100 001 eval (v16 ) = f (10 .5888 , 4.2146 ) = 34 .3595
v17 = 1110011001 1000010000 0101010111 011 eval (v17 ) = f (10 .5888 , 4.2425 ) = 32 .9321
v18 = 1110011001 1000010000 0101010111 001 eval (v18 ) = f (10 .5888 , 4.2424 ) = 32 .9567
v19 = 1111011000 1000101000 1110000010 001 eval (v19 ) = f (11 .5181, 4.4728 ) = 19 .6697
v20 = 1110011001 1000010000 0101010111 001 eval (v20 ) = f (10 .5888 , 4.2424 ) = 32 .9567
Total = 625 .3728
Hybrid intelligent systems

Evolutionary neural networks

 Evolutionary neural networks

■ Although neural networks are used for solving a

variety of problems, they still have some
limitations.
■ One of the most common is associated with neural
network training. The back-propagation learning
algorithm cannot guarantee an optimal solution.
In real-world applications, the back-propagation
algorithm might converge to a set of sub-optimal
weights from which it cannot escape. As a result,
the neural network is often unable to find a
desirable solution to a problem at hand.
 Evolutionary neural networks

■ Another difficulty is related to selecting an

optimal topology for the neural network. The
“right” network architecture for a particular
problem is often chosen by means of heuristics,
and designing a neural network topology is still
more art than engineering.
■ Genetic algorithms are an effective optimisation
technique that can guide both weight optimisation
and topology selection.
 Encoding a set of weights in a chromosome

x1 0.9
1 4
-0.8
From neuron: 1 2 3 4 5 6 7 8
0.1 -0.6
To neuron: 1 0 0 0 0 0 0 0 0 0.4
2 0 0 0 0 0 0 0 0 5
-0.3 0.1
3 0 0 0 0 0 0 0 0
x2 0.6 y
4 0.9 -0.3 -0.7 0 0 0 0 0 2 8
5 -0.8 0.6 0.3 0 0 0 0 0 -0.2
-0.2
0.5
6 0.1 -0.2 0.2 0 0 0 0 0 6
7 0.4 0.5 0.8 0 0 0 0 0 -0.7
0.3 0.9
8 0 0 0 -0.6 0.1 -0.2 0.9 0
0.2
x3
3 0.8 7

Chromosome: 0.9 -0.3 -0.7 -0.8 0.6 0.3 0.1 -0.2 0.2 0.4 0.5 0.8 -0.6 0.1 -0.2 0.9
 Fitness Function
■ The second step is to define a fitness function for
evaluating the chromosome’s performance. This
function must estimate the performance of a
given neural network. We can apply here a
simple function defined by the sum of squared
errors.
■ The training set of examples is presented to the
network, and the sum of squared errors is
calculated. The smaller the sum, the fitter the
chromosome. The genetic algorithm attempts
to find a set of weights that minimises the sum
of squared errors.
 weight optimisation
Crossover & Mutation
■ The third step is to choose the genetic operators –
crossover and mutation. A crossover operator
takes two parent chromosomes and creates a
single child with genetic material from both
parents. Each gene in the child’s chromosome is
represented by the corresponding gene of the
randomly selected parent.
■ A mutation operator selects a gene in a
chromosome and adds a small random value
between −1 and 1 to each weight in this gene.
 Crossover in weight optimisation
Parent 1 Parent 2
3 3
x1 -0.6 x1 0.3 0.3
0.1
1 1
-0.2 0.9
0.4 0.5
y -0.8 -0.9
y
-0.7
4 6 0.2
4 6
x2 0.9 x2 -0.5
2 -0.8 2 0.6
-0.3 -0.1

5 5
0.1 -0.7 -0.2 0.9 0.4 -0.3 -0.6 0.5 -0.8 0.3 0.2 0.9 -0.5 -0.8 -0.1 0.3 -0.9 0.6

0.1 -0.7 0.9 -0.5 -0.8 0.1 -0.6 0.5 -0.8

Child
3
x1 0.1 -0.6
1 0.9
-0.8 0.5
y
-0.7
4 6
x2 -0.5
2 -0.8
-0.1

5
 Mutation in weight optimisation

Original network Mutated network

3 3
x1 -0.6 x1 -0.6
0.1 0.1
1 1
-0.2 -0.2
0.4 0.5
y -0.1 0.5
y
-0.7
4 6 -0.7
4 6
x2 0.9 x2 0.9
2 -0.8 2 -0.8
-0.3 0.2

5 5
0.1 -0.7 -0.2 0.9 0.4 -0.3 -0.6 0.5 -0.8 0.1 -0.7 -0.2 0.9 -0.1 0.2 -0.6 0.5 -0.8
 Evolutionary neural networks:
optimal topology
Can genetic algorithms help us in selecting
the network architecture?
The architecture of the network (i.e. the number of
neurons and their interconnections) often
determines the success or failure of the application.
Usually the network architecture is decided by trial
and error; there is a great need for a method of
automatically designing the architecture for a
particular application. Genetic algorithms may
well be suited for this task.
 Evolutionary neural networks:
optimal topology

■ The basic idea behind evolving a suitable network

architecture is to conduct a genetic search in a
population of possible architectures.
■ We must first choose a method of encoding a
network’s architecture into a chromosome.
Encoding the network architecture
■ The connection topology of a neural network can
be represented by a square connectivity matrix.
■ Each entry in the matrix defines the type of
connection from one neuron (column) to another
(row), where 0 means no connection and 1
denotes connection for which the weight can be
changed through learning.
■ To transform the connectivity matrix into a
chromosome, we need only to string the rows of
the matrix together.
 Encoding of the network topology

3
From neuron: 1 2 3 4 5 6
x1
To neuron: 1 0 0 0 0 0 0 1
2 0 0 0 0 0 0 y
3 1 1 0 0 0 0 4 6
4 1 0 0 0 0 0 x2
5 0 1 0 0 0 0 2
6 0 1 1 1 1 0
5
Chromosome:
0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 1 1 1 1 0
The cycle of evolving a neural network topology
Neural Network j

Generation i
Fitness = 117

Training Data Set

0 0 1.0000
0.1000 0.0998 0.8869 Crossover
0.2000 0.1987 0.7551 Parent 1
Generation (i + 1) 0.3000 0.2955 0.6142
0.4000 0.3894 0.4720 Parent 2
0.5000 0.4794 0.3345
0.6000 0.5646 0.2060
0.7000 0.6442 0.0892 Child 1
0.8000 0.7174 -0.0143
0.9000 0.7833 -0.1038 Child 2
1.0000 0.8415 -0.1794

Mutation
Questions? Discussion?
Suggestions ?