| Class | Description |
|---|---|
| GGABinaryIncon |
Main class of Generational Genetic Algorithm for feature selection using inconsistency ratio as evaluation measure
This class implements a Generational GA with binary representation for feature selection Inconsistency ratio (FILTER)
Inputs: nearest neighbours for K-NN, k value for k-Tournamentvalor, crossover probability,
mutation probability, seed, population size, number of evaluations, alfa value for fitness balancing
Encoding: Binary
Selection: k-Tournament
Replacement: Descendants always replaces parents
Crossover and mutation in one point
Fitness: (1-alfa)*%hits + alfa*features_selected
Stopping criteria: number of evaluations
|
| Main |
File: Main.java
This is the main class of the algorithm.
|