Authors

  • Juraev Gulomjon Primovich
    Associate Professor of the Department of Methods of Exact and Natural Sciences, National Center for Teacher Training in New Methods of the Kashkadarya Region, Uzbekistan
  • Saparov Saidkul Khojamurodovich
    Doctoral student, Tashkent University of Information Technologies named after Muhammad al-Khwarizmi, Uzbekistan

DOI:

https://doi.org/10.71337/inlibrary.uz.ijasr.131676

Keywords:

Primary data processing training sample classification

Abstract

In the primary processing of information, in particular in character recognition, an important issue is the selection and classification of an informative feature or a set of features classifying objects. Despite the fact that a number of methods and algorithms have been proposed to solve these problems, there are many problems in this direction that are waiting to be solved. This is due to the fact that many of the proposed approaches strongly depend on the nature of the object of study, the number of its features, the type of perceived values of features, the size of the study sample, etc., and impose certain requirements on the above. In addition, each method or algorithm will strongly depend on whether the criterion of informative selection of features and the defining rule determining the quality of the choice made are correctly chosen.

This article presents a description of the algorithm developed taking into account the above approaches to the selection of information complexes of signs, as well as recommendations on the application of this algorithm in practical matters of the medical field, i.e. in ischemic heart disease obtained as an object of study (5 classes, 507 objects, 89 signs, including X_1 class “strenuous angina”, X_2  class “Acute myocardial infarction”, class X_3 “Arrhythmic form”, class X_4 “Postinfarction cardiosclerosis”, for class X_5 “Persistent form of atrial fibrillation”) formulated training was applied to the selection and positive results were achieved.


background image

Volume 04 Issue 03-2024

149



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135




















































A

BSTRACT

In the primary processing of information, in particular in character recognition, an important issue is the
selection and classification of an informative feature or a set of features classifying objects. Despite the fact
that a number of methods and algorithms have been proposed to solve these problems, there are many
problems in this direction that are waiting to be solved. This is due to the fact that many of the proposed
approaches strongly depend on the nature of the object of study, the number of its features, the type of
perceived values of features, the size of the study sample, etc., and impose certain requirements on the
above. In addition, each method or algorithm will strongly depend on whether the criterion of informative
selection of features and the defining rule determining the quality of the choice made are correctly chosen.

This article presents a description of the algorithm developed taking into account the above approaches to
the selection of information complexes of signs, as well as recommendations on the application of this
algorithm in practical matters of the medical field, i.e. in ischemic heart disease obtained as an object of

study (5 classes, 507 objects, 89 signs, including X_1 class “strenuous angina”, X_2 class “Acute myocardi

al

infarction”, class X_3 “Arrhythmic form”, class X_4 “Postinfarction cardiosclerosis”, for class X_5 “Persistent

Journal

Website:

http://sciencebring.co
m/index.php/ijasr

Copyright:

Original

content from this work
may be used under the
terms of the creative
commons

attributes

4.0 licence.

Research Article

AN ALGORITHM FOR SELECTING AN INFORMATIVE
SYMBOLIC COMPLEX BASED ON CLASSIFICATION ERROR
COEFFICIENTS AND PROBABILISTIC INDICATORS IN THE
REPRESENTATION OF SYMBOLS


Submission Date:

March 20,

2024,

Accepted Date:

March 25, 2024,

Published Date:

March 30, 2024

Crossref doi:

https://doi.org/10.37547/ijasr-04-03-27


Juraev Gulomjon Primovich

Associate Professor of the Department of Methods of Exact and Natural Sciences, National Center for
Teacher Training in New Methods of the Kashkadarya Region, Uzbekistan

Saparov Saidkul Khojamurodovich

Doctoral student, Tashkent University of Information Technologies named after Muhammad al-Khwarizmi,
Uzbekistan


background image

Volume 04 Issue 03-2024

150



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































form of atrial fibrillation”) formulated training was applied to the selection and positive results were

achieved.

K

EYWORDS

Primary data processing, training sample, classification, selection of an informative symbolic complex,
classification error.

I

NTRODUCTION

The experience conducted during the pandemic
in the world has shown the need to accelerate the
work on digitalization of medicine while
eliminating problems in the field of medicine. In
particular, in the field of medicine, special
attention is also paid to the development and
development of data mining methods that
correspond to human decisions.
Important scientific projects carried out by
scientific societies and research institutes
engaged in medical research around the world
are genomics and the Human Genome project,
vaccines and vaccination, medical robots, blood
detection and analysis technologies, medical
digital technologies, geonomics (bioanalysis),
medical rehabilitation technologies, the latest
scientific research in the field of medicine and the
results count. These researches, aimed at the
scientific development of medical science, serve
as the basis for the intellectual analysis of medical
information in solving classification issues,
choosing a set of informative signs, forming
symptom complexes and diagnosing diseases
based on them. Usually, the size of the medical
information that the disease evaluator collects
and stores is tens or hundreds of characters.
Based on this information, the decision-making

process by industry professionals becomes an
extremely complex process.
Therefore, issues related to the transition from a
large size to a smaller one, which is more
important in the primary processing of medical
information, that is, with the selection and
evaluation of information sign complexes, the
formation of symptom complexes in the context
of information sign complexes and the
development of modern recognition systems that
help industry specialists in solving issues such as
diagnosis on based on them. In particular, the
decision on the transition from a large size to a
smaller significant size during primary data
processing, that is, on the set of informative
symbolic complexes, was of interest to scientists
around the world, and in this regard, a number of
studies are underway [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
11, 12, 13, 14, 15].
This article also develops an algorithm for
selecting an informative set of symbols based on
classification error coefficients and probabilistic
indicators in the representation of symbols.

T

HE MAIN PART

The standard was chosen for the preparation of

𝑥

𝑝1

, 𝑥

𝑝2

, … , 𝑥

𝑝𝑚

𝑝

∈ 𝑋

𝑝

,

𝑝 = 1, 𝑟;

̅̅̅̅̅

cognitively, the


background image

Volume 04 Issue 03-2024

151



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































question of choosing a set of characters is the
symbol value with a space in the part

𝜆 =

(𝜆

1

, 𝜆

2

, … , 𝜆

𝑁

), 𝜆

𝑗

∈ {0,1}, 𝑗 = 1, 𝑁;

̅̅̅̅̅̅

is included in

the vector.
Here, the main problem of the

𝜆

vector is to

ensure the transition from an N-dimensional
character space to an

-dimensional character

space small enough for the

𝑁

number.

Here

𝜆

means that the signs corresponding to the

components of the vector that are equivalent
together mean that the selected part has a set of
characters in space, and signs equal to zero mean
that the corresponding characters do not
participate in the extracted information character
set.

Definition 1.

The studied vector space

𝜆

is called

-dimensional if, for an arbitrary vector

𝜆

under

consideration, the sum of its components is equal
to

∑ 𝜆

𝑗

𝑁

𝑗=1

= ℓ

Similarly, in an

-dimensional vector space, let

the set in which the vectors

𝜆

are located be

denoted by

Λ

.

By definition, the mathematical expression

of this set will be:

Λ

= { 𝜆: ∑ 𝜆

𝑗

𝑁

𝑗=1

= ℓ , 𝜆

𝑗

∈ {0,1}, 𝑗 = 1, 𝑁;

̅̅̅̅̅̅ }

here, set of

Λ

number of vectors

𝜆

is equal to

𝐶

𝑁

=

𝑁!

ℓ!(𝑁−ℓ)!

.

if

ℓ = 1, 𝑁

̅̅̅̅̅

, in this set

2

𝑁

− 1

is

𝜆

vectors:

∑ 𝐶

𝑁

=

𝑁

ℓ=1

2

𝑁

− 1.

Definition 2.

Λ

vectors

𝜆

which are

elements of the set

are called informative

vectors. The number of non-zero components of
these vectors is

. Suppose that the objects of the

training sample based on the reference table

𝜃(𝑁)

have a classification error, and the number of
incorrectly recognized objects is

ϰ(𝑁)

.

Learning here on the other hand, the total number
of objects in the sample is

𝑀

, then the

classification error in

𝑁

-dimensional space is

calculated using the formula

𝜃(𝑁) =

𝜘(𝑁)

𝑀

.

Let the mechanism of operation of the

proposed algorithm be defined as follows. First,
the probability vector for selecting symbols of the
objects

of

the

training

sample

𝑝

𝜈

=

(𝑝

𝜈

1

, 𝑝

𝜈

2

, … , 𝑝

𝜈

𝑁

)

is introduced.

Here

𝑝

𝜈

means that the

𝜈

in the index of the

probability vector is the probability vector.
Usually, this index will be associated with the
choice of a probability vector. The first choice of

𝜈 = 1

bo‘lib,

𝑝

1

𝑗

=

1

𝑁

; 𝑗 = 1, 𝑁

̅̅̅̅̅;

Similarly,

𝑝

𝜈

= (𝑝

𝜈

1

, 𝑝

𝜈

2

, … , 𝑝

𝜈

𝑁

)

probability

vector

is informative

𝜆 ∈ Λ

in vector space we

want

𝑝

𝜈

|

𝜆

= (𝜆

1

𝑝

𝜈

1

,

𝜆

2

𝑝

𝜈

2

, … , 𝜆

𝑁

𝑝

𝜈

𝑁

)

the form is

characterized.

Based on the initial given probability,

Λ

from the set

𝑝

𝜈

=

(𝑝

𝜈

1

, 𝑝

𝜈

2

, … , 𝑝

𝜈

𝑁

)

, to the

probability vector

𝜆

, the vector is randomly

selected, and this is denoted as

𝜆

𝜈1

, 𝜆

𝜈2

, … , 𝜆

𝜈𝑘

. .


In the context of all the features of this

training sample, the classification process is


background image

Volume 04 Issue 03-2024

152



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































carried out on the basis of sequential calculation
of the following Formulas (1), (2) and (3:

1.

Initially, each object belonging to the

𝑋

𝑝

class

is compared with objects in its class, as well
as with objects in other classes

𝑥

𝑝𝑖

the

proximity function between the object and
objects

𝑥

𝑝𝑞

of the

𝑋

𝑝

class

𝜌

𝑗

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

)

is

calculated as:

𝜌

𝑗

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

)

= {

1, if |𝑥

𝑝𝑖

𝑗

− 𝑥

𝑝𝑞

𝑗

| = 0

0, otherwise

(1)

Here

𝑗 = 1, 𝑁;

̅̅̅̅̅̅ 𝑖𝑠 𝑒𝑞𝑢𝑎𝑙 𝑖, 𝑞 = 1, 𝑚

𝑝

̅̅̅̅̅̅̅; 𝑖 ≠ 𝑞;

.

2. Comparative evaluation for each class,

that is, each object of this class

𝑋

𝑝

is

Г

𝑝

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

)

for

𝑥

𝑝𝑖

Г

𝑝

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

) =

1

𝑚

𝑝

∑ ∑ 𝜌

𝑗

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

), 𝑖 = 1, 𝑚

𝑝

; 𝑖

𝑁

𝑗=1

𝑚

𝑝

𝑞=1

≠ 𝑞 ; (2)

3.

Based on the results obtained in a

comparative assessment, the maximization

problem is solved using the formula


Г

𝑝

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

) = max

𝑝=1,𝑟;

̅̅̅̅̅

Г

𝑝

(𝑥

𝑝𝑖,

𝑥

𝑝𝑞

), 𝑖, 𝑞 = 1, 𝑚

𝑝

; 𝑖

≠ 𝑞; (3)

And in result

𝑥

𝑝𝑖

∈ 𝑋

𝑝

.

It determines the levels of individual

significance of all features in the reference sample
of training. This is a reference training without
selection

𝑝

𝜈

=

(𝑝

𝜈

1

, 𝑝

𝜈

2

, … , 𝑝

𝜈

𝑁

)

using

the

probability vector, the column containing one

character is excluded from the training sample
and the classification error coefficient is
calculated on the segment of the remaining
characters

𝜃(𝑁 −

1)

. These processes include

the following two important cases.

The first case:

the difference in

classification errors if

𝜃(𝑁) =

𝜃(𝑁 − 1)

, in this

case, this character is completely removed from
the training sample, and the number of characters
is reduced by one character.

The second case:

otherwise

,

𝜃(𝑁) ≠

𝜃(𝑁 − 1)

, in this case, this symbol will be left in

the training sample, and the process will start
againIn the second case, the number of signs of
educational selection does not change.

Description of the proposed algorithm

1-step.

Initial classification error is

𝜃(𝑁)

;

2-step.

the value

is entered ,

ℓ << 𝑁

;

3-step. In

𝜈 = 1

is

𝑝

𝜈

= (𝑝

𝜈

1

, 𝑝

𝜈

2

, … , 𝑝

𝜈

𝑁

)

for

𝑝

𝜈

𝑗

=

1

𝑁

; 𝑗 = 1, 𝑁;

̅̅̅̅̅̅

initial values are assigned;

4-step.

One character is randomly selected

from

𝑁

characters and subtracted from the array.

Than, In the cross section of

𝑁 − 1

characters, the

classification process is performed. In a result for

𝑁 − 1

characters, the error coefficient is equal to

𝜃(𝑁 − 1)

;

5-step.

If

𝜃(𝑁) = 𝜃(𝑁 − 1)

, here, in this

case, this symbol is completely removed from the
training sample, otherwise this symbol remains in
the training sample, and the process begins with
step 4. This process is reversible as long as

𝑁 −

ℎ = ℓ

.

If

𝑁 − ℎ = ℓ

, in this case is formed number

of the signs

and a signal solution with multiple


background image

Volume 04 Issue 03-2024

153



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































characters is formulated and proceeds to step 6.
Similarly, if equality is not satisfied, the process
continues until all characters are processed one
by one. However, after all the symbols are
considered, we proceed to the next step;

6-step.

Suppose that

𝑁 − ℎ = ℓ

, then given

the

following

𝑝

𝑗

=

1

2𝑁−ℎ

, 𝑗 =

1, ℎ;

̅̅̅̅̅̅ 𝑝

𝑗

=

2

2𝑁−ℎ

, 𝑗 = ℎ + 1, 𝑁;

̅̅̅̅̅̅̅̅̅̅̅̅

with probability 4

process,

in the reverse step, only those symbols that were
selected earlier, which should be chosen

randomly, and those that remained in the system
differ from each other in probabilities.

Then go to

step 5.

With the help of a software package created based
on this algorithm, informative symbolic medicine
complexes for coronary heart disease were
selected (see Table 1).
1-table

information symbolic complexes selected in
103 variants when

𝓵 = 𝟗

𝓵 = 𝟗

𝓵 = 𝟗

𝓵 = 𝟗

𝓵 = 𝟗

𝑥

3

, 𝑥

5

, 𝑥

10

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

56

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

48

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

64

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

63

, 𝑥

72

, 𝑥

80

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

49

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

10

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

68

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

14

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

71

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

85

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

34

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

48

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

25

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

64

, 𝑥

80

, 𝑥

89

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

81

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

32

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

25

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

10

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

24

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

42

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

60

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

14

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

38

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

42

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

21

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

19

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

68

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

77

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

24

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

72

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

24

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

24

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

85

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

71

, 𝑥

80

𝑥

3

, 𝑥

4

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

12

, 𝑥

24

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

42

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

72

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

26

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

48

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

49

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

71

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

38

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

38

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

.

𝑥

3

, 𝑥

11

, 𝑥

12

, 𝑥

24

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

26

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

20

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

38

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

38

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

19

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

77

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

54

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

38

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

54

, 𝑥

62

, 𝑥

72

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

64

, 𝑥

80

, 𝑥

85

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

, 𝑥

85

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

42

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

18

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89


background image

Volume 04 Issue 03-2024

154



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

32

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

81

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

26

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

36

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

, 𝑥

85

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

22

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

60

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

45

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

10

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

64

, 𝑥

80

, 𝑥

85

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

22

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

35

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

68

, 𝑥

80

, 𝑥

89

𝑥

10

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

64

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

4

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

52

, 𝑥

53

, 𝑥

61

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

49

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

60

, 𝑥

80

𝑥

3

, 𝑥

11

, 𝑥

19

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

63

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

36

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

49

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

29

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

25

, 𝑥

40

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

29

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

68

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

41

, 𝑥

48

, 𝑥

52

, 𝑥

53

, 𝑥

80

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

68

, 𝑥

80

𝑥

3

, 𝑥

10

, 𝑥

11

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

.

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

52

, 𝑥

53

, 𝑥

60

, 𝑥

62

, 𝑥

80

.

𝑥

3

, 𝑥

11

, 𝑥

40

, 𝑥

42

, 𝑥

52

, 𝑥

53

, 𝑥

59

, 𝑥

62

, 𝑥

80

.

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

22

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

63

, 𝑥

80

.

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

40

, 𝑥

48

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

.

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

45

, 𝑥

52

, 𝑥

53

, 𝑥

62

, 𝑥

80

, 𝑥

89

.

𝑥

3

, 𝑥

5

, 𝑥

11

, 𝑥

20

, 𝑥

41

, 𝑥

52

, 𝑥

53

, 𝑥

80

, 𝑥

89

.

C

ONCLUSION

As a criterion of informativeness in solving the
problem of choosing a set of informative features,
which is one of the main tasks of primary data
processing, the classification error coefficient and
the probability vector of the features of the
sample objects were obtained, aimed at reducing
the classification error. In addition, the
probability vector used when selecting symbols
prevents the inappropriate exclusion of
significant object symbols from the selection.

With the help of a software package developed on
the basis of the proposed algorithm, it was
possible to select and diagnose (classify) based on
a set of informative signs that are part of coronary
heart disease, which can serve as a basis for the
diagnosis of the disease, that is, the class is

considered as class

𝑋

1

-

“Strenuous angina”, class

𝑋

2

“Acute myocardial infarction”, class

𝑋

3

“Arrhythmic form”, class

𝑋

4

“Postinfarction

cardiosclerosis”, class

𝑋

5

“Persistent form of atrial

fibrillation, in particular, based on clinical signs in
patients.

R

EFERENCES

1.

Bykova V.V., Kataeva A.V. Methods and means
of analyzing the informative value of signs in
the processing of medical data//Software

products and systems /Software & Systems №

2 (114), 2016.-c. 172-178.

2.

Fazylov Sh.Kh., Nishanov A.Kh., Mamatov N.S.
Methods and algorithms for selecting
informative features based on heuristic
criteria of informativeness//Monograph.-T.:
Fan wa technology.-Tashkent, 2017.-132 p.


background image

Volume 04 Issue 03-2024

155



International Journal of Advance Scientific Research
(ISSN

2750-1396)

VOLUME

04

ISSUE

03

Pages:

149-155

SJIF

I

MPACT

FACTOR

(2022:

5.636

)

(2023:

6.741

)

(2024:

7.874

)

OCLC

1368736135















































3.

Ashok B., Aruna P. Comparison of Feature
selection methods for diagnosis of cervical
cancer using SVM classifier//Journal of
Engineering Research and Applications. ISSN:
2248-9622, Vol. 6, Issue 1, (Part-1) January
2016.-pp. 94-99.

4.

Bolón-Canedo, V. & Alonso-betanzos, A.
Ensembles for feature selection: A review and
future

trends//Information

Fusion

52(2019).-

рр. 1

-12.

5.

Emary, E., Zawbaa, H. M. & Hassanien, A. E.
Binary grey wolf optimization approaches for
feature selection//Neurocomputing 172,
2016.-

рр.371

-381.

6.

Faris, H. et al. An efficient binary Salp Swarm
Algorithm with crossover scheme for feature
selection

problems//Knowledge-based

Systems 154, 2018.-pp. 43-67.

7.

Gao, W., Hu, L. & Zhang, P. Class-specific
mutual information variation for feature
selection//Pattern Recognition 79, 2018.-pp.
328-339.

8.

Gao, W., Hu, L., Zhang, P. & He, J. Feature
selection considering the composition of
feature

relevancy//Pattern

Recognition

Letters 112, 2018.-pp. 70-74.

9.

Hussien A., Hassanien A., Houssein E., et al.See
more. S-shaped binary whale optimization
algorithm for feature selection//Advances in
Intelligent Systems and Computing, Vol. 727,
2019.-pp. 79-87.

10.

Li, J. & Liu, H. Challenges of Feature Selection
for Big Data Analytics//IEEE Intelligent
Systems 32, (2017).-

рр. 9

-15.

11.

Liu, C., Wang, W., Zhao, Q., Shen, X. & Konan, M.
A new feature selection method based on a

validity index of feature subset. Pattern
Recognition Letters 92, (2017).-

рр. 1

-8.

12.

Nishanov А.Kh., Djurayev G.P., Kasanova М.K

h.

Improved

algorithms

for

calculating

evaluations

in

processing

medical

data//Compusoft: An International Journal of
Advanced Computer Technology, 8(6), June-
2019.-pp. 3158-3165.

13.

Nishanov A.Kh., Akbaraliev B.B., Juraev G.P.,
Khasanova M.A., Maksudova M.Kh., Umarova
Z.F. The algorithm for selection of symptom
complex of ischemic heart diseases based on
flexible search//Journal of Cardiovascular
Disease Research, Vol. 11(2), 2020.-pp. 218-
223.

14.

Nishanov A.K., Akbaraliev B.B. and Djurayev
G.P., A Symptom Selection Algorithm Based on
Classification

Errors//International

Conference on Information Science and
Communications

Technologies

(ICISCT),

2020.-pp. 1-4.

15.

Nishanov A.Kh., Djurayev, G.P., Khasanova,
M.A. Classification and feature selection in
medical data preprocessing//Compusoft: An
International Journal of Advanced Computer
Technology, 9(6), June-2020.-pp. 3725-3732.

References

Bykova V.V., Kataeva A.V. Methods and means of analyzing the informative value of signs in the processing of medical data//Software products and systems /Software & Systems № 2 (114), 2016.-c. 172-178.

Fazylov Sh.Kh., Nishanov A.Kh., Mamatov N.S. Methods and algorithms for selecting informative features based on heuristic criteria of informativeness//Monograph.-T.: Fan wa technology.-Tashkent, 2017.-132 p.

Ashok B., Aruna P. Comparison of Feature selection methods for diagnosis of cervical cancer using SVM classifier//Journal of Engineering Research and Applications. ISSN: 2248-9622, Vol. 6, Issue 1, (Part-1) January 2016.-pp. 94-99.

Bolón-Canedo, V. & Alonso-betanzos, A. Ensembles for feature selection: A review and future trends//Information Fusion 52(2019).-рр. 1-12.

Emary, E., Zawbaa, H. M. & Hassanien, A. E. Binary grey wolf optimization approaches for feature selection//Neurocomputing 172, 2016.-рр.371-381.

Faris, H. et al. An efficient binary Salp Swarm Algorithm with crossover scheme for feature selection problems//Knowledge-based Systems 154, 2018.-pp. 43-67.

Gao, W., Hu, L. & Zhang, P. Class-specific mutual information variation for feature selection//Pattern Recognition 79, 2018.-pp. 328-339.

Gao, W., Hu, L., Zhang, P. & He, J. Feature selection considering the composition of feature relevancy//Pattern Recognition Letters 112, 2018.-pp. 70-74.

Hussien A., Hassanien A., Houssein E., et al.See more. S-shaped binary whale optimization algorithm for feature selection//Advances in Intelligent Systems and Computing, Vol. 727, 2019.-pp. 79-87.

Li, J. & Liu, H. Challenges of Feature Selection for Big Data Analytics//IEEE Intelligent Systems 32, (2017).-рр. 9-15.

Liu, C., Wang, W., Zhao, Q., Shen, X. & Konan, M. A new feature selection method based on a validity index of feature subset. Pattern Recognition Letters 92, (2017).-рр. 1-8.

Nishanov А.Kh., Djurayev G.P., Kasanova М.Kh. Improved algorithms for calculating evaluations in processing medical data//Compusoft: An International Journal of Advanced Computer Technology, 8(6), June-2019.-pp. 3158-3165.

Nishanov A.Kh., Akbaraliev B.B., Juraev G.P., Khasanova M.A., Maksudova M.Kh., Umarova Z.F. The algorithm for selection of symptom complex of ischemic heart diseases based on flexible search//Journal of Cardiovascular Disease Research, Vol. 11(2), 2020.-pp. 218-223.

Nishanov A.K., Akbaraliev B.B. and Djurayev G.P., A Symptom Selection Algorithm Based on Classification Errors//International Conference on Information Science and Communications Technologies (ICISCT), 2020.-pp. 1-4.

Nishanov A.Kh., Djurayev, G.P., Khasanova, M.A. Classification and feature selection in medical data preprocessing//Compusoft: An International Journal of Advanced Computer Technology, 9(6), June-2020.-pp. 3725-3732.