Comparative Analysis of Association Rule Mining Algorithms Breast Cancer Diagnosis
This paper presents Frequent items generation in popular data mining technique with the classic apriori algorithm along with a modified version aforementioned ECLAT within a constant data size but differently arranged data. Application of such algorithms in determining the efficiency of the algorithm on a diagnostic breast cancer data which consist of feature information computed from images in digital form from fine needle aspirate(FNA) of a breast mass based on features radius, texture, fractal dimension, perimeter, area, smoothness, concavity their mean and their worst values. Also it gives Comparative analysis of algorithms apriori and Equivalence Class Clustering and Bootom up Lattice Traversal (ECLAT) in an automatic diagnosis system for finding association rules in diagnostic breast cancer data.
Keywords - Apriori, Eclat, association rules, Pre-processing