|
Activity Number:
|
272
- Approaches in Clustering for Analysis of Emerging Data Types
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 9, 2022 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
|
Abstract #322633
|
|
|
Title:
|
Mixtures of Matrix Variate Contaminated Normal Distributions
|
|
Author(s):
|
Salvatore Daniele Tomarchio and Michael Gallaugher* and Antonio Punzo and Paul David McNicholas
|
|
Companies:
|
University of Catania and Baylor University and University of Catania and McMaster University
|
|
Keywords:
|
mixture model;
outlier detection;
matrix variate distributions;
three-way data
|
|
Abstract:
|
Clustering and classification is the process of finding and analyzing underlying homogenous group structure in heterogenous data. In a cluster analysis the ability to model as well as detect outlying observations is an important task. In the case of higher order data that come in the form of matrices or tensors, the detection of outliers becomes even more difficult and needs to be performed via numerical and probabilistic methods. In this presentation, a mixture of contaminated matrix variate normal distributions will be presented for the detection of outlying matrices in a cluster analysis. Simulated and real data will be used for analysis.
|
Authors who are presenting talks have a * after their name.
