Transcript Step 1

An Approach of Artificial
Intelligence Application for
Laboratory Tests Evaluation
Ş.l.univ.dr.ing. Corina SĂVULESCU
University of Piteşti
The principal domains where GA
have successfully applied to
optimization problems





function optimization
image processing
classification and machine learning
training of neural networks
systems’ control
Why using a GA?





are stochastic algorithms
use a vocabulary borrowed from natural
genetics
are more robust than existing directed search
methods
maintain a population of potential solutions
the structure of a simple GA is the same as
the structure of any evolution program
A GA for a particular problem must
have the following five components:





a genetic representation for potential
solutions to the problem
a way to create an initial population of
potential solutions
an evaluation function that plays the role
of environment rating solution in term of
their “fitness”
a genetic operator that alter composition
of children
a set of values for various parameters that
the genetic algorithm uses
GA’s principles
N individuals
N individuals
N individuals
N individuals
Fitness
Generation 3
Generation 2
Generation 1
Generation 0
The structure of the chosen
genetic algorithm
Step 1:
Generation of initial
population P(t)
The structure of the chosen
genetic algorithm
n
 f (x )  y
i 1
i
f  1/ S
i
The structure of the chosen
genetic algorithm
Step 3:
The population's chromosomes are
sorted based on their fitness value
determined during the previous step
The structure of the chosen
genetic algorithm
Step 4:
The best chromosomes are selected, and
they will be placed unconditionally in the
next population P(t+1)
5%
50 %
30 %
15 %
The structure of the chosen
genetic algorithm
Step 5:
The chromosomes that are object to the
crossover operator are then selected
1
2/3
1/3
8x
3/2
5/6
2/3
5/6
13/6
N=8
The structure of the chosen
genetic algorithm
Step 6:
The descendants from the previous
step are subject to the mutation
operator, resulting new members for
the P(t+1) population
The structure of the chosen
genetic algorithm
Step 7:
The population P(t+1) is completed with
individuals selected randomly from the
P(t) population
The application description
Fig. 1 – System's index response
Results of the system
identification
ξ
ω
n
(rad/sec)
Original model
0.6
2.5
Model identified
without noise
0.61
2.59
Model identified
with noise
0.65
2.79
Where ξ ω
n
are the function’s parameters:
y (t )  1 
  t
n
e
sin( t 1   2   )
n
1 2
Identified system's response
The application of the genetic
algorithm in electrophoresis tests
Positioning the agarose gel
The application of serum on the
agarose gel
The electrophoresis machine
Drying incubator
An example of results using the
agarose gel
The applications of GA to the
electrophoresis tests
Application of the genetic
algorithm in electrophoresis tests
The results obtained from using
a GA from the same example
The results obtained from using
a GA from the same example
The test result
Conclusions


This application is an alternative method
for evaluation of the laboratory tests (in
special electrophoresis tests), using
artificial intelligence.
The main advantage of this method is
the need of minimal medical knowledge.
Therefore, GA implementation is an
instrument easy to use by low/medium
trained personnel, offering tests results
quickly and clearly.