Online Program Home
My Program

Abstract Details

Activity Number: 182 - SPEED: New Methods in Statistical Genomics and Genetics Part 2
Type: Contributed
Date/Time: Monday, July 29, 2019 : 10:30 AM to 11:15 AM
Sponsor: Section on Statistics in Genomics and Genetics
Abstract #307555
Title: OASW Clustering
Author(s): Fatima Batool*
Keywords: Data clustering; clustering heuristic; clustering quality measure; average silhouette width; algorithm initialisation ; estimation of number of clusters

The two major problems in cluster analysis are: how many clusters are present in the data? And how to find the actual clustering solution? This study aims at developing a unified approach to estimate number of clusters and finding clustering solution mutually. In this work we adopt a simple yet powerful approach. We have proposed clustering methodology by optimizing average silhouette width (ASW) clustering quality index proposed by Rousseeuw in 1987 that can itself estimate number of clusters on the fly as well as produce the clustering against this estimated number. The proposed clustering algorithm requires just the data or pairwise distances for clustering and no other parameters are required to specify as an input. Extensive simulations were conducted for a vast range of challenges to characterize the proposed method and to evaluate the performance against competitors. One of the major results of the simulations was the ASW based clustering optimizer functions can produce desirable clusters but might not estimate that number of clusters. Which is a surprising result because ASW is in extensive use for the estimation of number of clusters rather than for finding clustering.

Authors who are presenting talks have a * after their name.

Back to the full JSM 2019 program