EVOLUTION OF NEURAL NETWORK AND POLYNOMIAL NETWORK - Knowledge Mining Using Intelligent Agents - page 76

Table 3.2. Divison of dataset and its pattern distribution.

Datasets Patterns Patterns Patterns Patterns

in class 1 in class 2 in class 3

AUSTRA
Set1 345 153 192 —

Set2 345 154 191 —

BALANCE
Set1 313 144 25 144

Set2 312 144 24 144

CREDIT
Set1

500 350 150 —

Set2 500 350 150 —

HABERMAN
Set1 153 112 41 —

Set2 153 113 40 —

HOUSE
Set1 217 133 84 —

Set2 218 134 84 —

IRIS
Set1 75 25 25 25

Set2 75 25 25 25

MONK
Set1 62 31 31 —

Set2 62 31 31 —

PIMA
Set1 384 134 250 —

Set2 384 134 250 —

WBC
Set1 350 229 121 —

Set2 349 229 120 —

WINE
Set1 89 29 36 24

Set2 89 30 35 24

Table 3.3. Parameters of PSO considered for simulation.

Parameters Values

Population Size 30 Maximum Iterations 1000

Inertia Weight 0.729844

Cognitive Parameter 1.49445

Social Parameter 1.49445

Constriction Factor 1.0