Identification of AIDS Disease Severity Using Genetic Algorithm - Computational Intelligence Techniques for Comparative Genomics - page 98

Table 2 AIDS patient Dataset
S.No. Age Gender Marital status CD4 cell count

(cells/mm 3 )

AIDS

1 21 Male Unmarried 450 Positive (AIDS)
2 18 Male Unmarried 650 Negative (No
AIDS)
3 16 Female Unmarried 350 Positive
4 35 Female Married 202 Positive
5 25 Female Married 400 Positive
6 22 Male Married 150 Positive
7 29 Male Married 100 Positive
8 10 Male Unmarried 275 Positive
9 15 Female Unmarried 350 Positive
10 65 Female Married 65 Positive
11 50 Male Married 120 Positive
12 45 Female Married 185 Positive
13 12 Male Unmarried 405 Positive
14 08 Female Unmarried 375 Positive
15 27 Male Married 750 Negative
16 23 Female Married 850 Negative
17 32 Male Married 750 Negative
18 19 Male Unmarried 425 Positive
19 20 Female Unmarried 550 Positive
20 55 Male Married 55 Positive
AIDS patient dataset description is described using the following attributes: Age (children to old
age persons), Gender (Male and Female), Marital Status (Married and Unmarried), and CD4 Cell
Count is given in Table 1 and classes are AIDS is divided two types classes used on positive
(AIDS) and Negative (No AIDS). Those are following developed [ 12 ]

where A i represents the i-th predictor attribute,
Op i comparison operator {<, >, ≤ , ≥ , =} and V ij denote j-th value of the i-th

attribute (See Figs. 1 , 2 , and 3 ).
The absence of the attribute in the genome is represented with

symbol which
indicates the absence of the value in the rule. A central mechanism in the GA is the

“ # ”

tness function that plays vital role in optimizing a speci c problem.

Classi cation rule is in the form of: P to Q

Q Value is the number of samples in the dataset that are satisfying ante-
cedent and consequent in the rule.

P

˄

…………… A k Op k V kj

A i Op i V ij

Fig. 1 Individual genome representation