Research Article

An Adaptive Multiobjective Genetic Algorithm with Fuzzy -Means for Automatic Data Clustering

Algorithm 1

ADNSGA2-FCM.
input: Dataset
Initialize parameters FCM and NSGA-II including population size Pop,
itermax, , , , , Tmax.
Random to select initial number of clusters and random to generate initial cluster
centers to create a initialize population .
Decode each individual to obtain the cluster centers, and calculate the membership
degrees using Eq. (3).
Calculate new cluster centers of each individual using Eq. (4) based on and
Calculate the of each individual using Eq. (2) based on and
Calculate fitness values and of each individual using Eq. (10) and (11). Calculate
, store and at each iteration.
Non-dominated sorting and crowding distance operation for population.
Using the crowded comparison operator to select.
Calculate and using Eq. (16).
Generate offspring using genetic operation.
Recombination current generation and offspring to select next generation using
elitism operation.
Using majority voting technique to determine the number of cluster .
If the number satisfies the selection condition, go to step (14); else go to step (3).
Find all chromosomes whose cluster numbers are equal to from the population in
step (11). The new population is composed of these chromosomes.
Decode each individual to obtain the cluster centers, and calculate the membership
degrees using Eq. (3).
Calculate new cluster centers of each individual using Eq. (4) based on and
Calculate the of each individual using Eq. (2) based on and
Calculate fitness values and of each individual using Eq. (10) and (11).
Calculate , store and at each iteration.
Non-dominated sorting and crowding distance operation for population.
Using the crowded comparison operator to select.
Calculate and using Eq. (16).
Generate offspring using genetic operation.
Recombination current generation and offspring to select next generation using
elitism operation.
If ADNSGA2-FCM has not met the stopping criterion (
and ), else and go to step (14).
ā€‰ā€‰Return the best individual ().