Alternative Technique Reducing Complexity of Maximum Attribute Relation
Abstract: Clustering refers to
the method grouping the large data into the smaller groups based on the similarity
measure. Clustering techniques have been applied on numerical, categorical and
mix data. One of the categorical data clustering technique based on the soft
set theory is Maximum Attribute Relation (MAR). The MAR technique allows
calculating all of pair multi soft set made. However, the computational complexity
is still an issue of the technique. To overcome the drawback, the paper
proposes the alternative algorithm to decrease the complexity so get the faster
response time. In this paper, to get the similar results as MAR without
calculating all pair of soft set is proved. The alternative algorithm is
implemented in MATLAB Software, and then experimental is run on the 10
benchmark datasets. The results show that the alternative algorithm improves
the computational complexity in term of response time up to 36.46%.
Author: Iwan Tri Riyadi Yanto,
Imam Azhari
Journal Code: jptkomputergg150171
