Online Program Home
My Program

Abstract Details

Activity Number: 531 - SPEED: Statistical Computing: Methods, Implementation, and Application, Part 2
Type: Contributed
Date/Time: Wednesday, July 31, 2019 : 11:35 AM to 12:20 PM
Sponsor: Section on Statistical Computing
Abstract #307957
Title: Clustering Smoothed Dissimilarities in Tertiary Data: a Shrinkage-Based Approach
Author(s): Bridget Manning* and David Hitchcock
Companies: University of South Carolina and University of South Carolina
Keywords: clustering ; tertiary ; dissimilarity; classification ; pre-smoothing; shrinkage

Clustering is an important multivariate method whose goal is to partition objects into homogeneous groups. Several popular methods of clustering are based on the use of pairwise dissimilarities. However, noisy data can result in dissimilarities that do not reflect the clustering structure and yield partitions that may be less than ideal. To remedy this, we consider smoothing dissimilarities before clustering. We apply this method to tertiary data using a shrinkage-based method of estimating cell probabilities. We show, via simulation and an application using voting data, that pre-smoothing dissimilarities in situations with noisy data results in better partitioning results than those produced by traditional clustering methods.

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

Back to the full JSM 2019 program