|
Activity Number:
|
178
- Novel Applications and Extensions of Dimension Reduction Methods
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 29, 2019 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Statistical Learning and Data Science
|
|
Abstract #305282
|
Presentation
|
|
Title:
|
Graph-Based Dependency Criterion with Applications in Biology
|
|
Author(s):
|
Salimeh Yasaei Sekeh* and Alfred O. Hero
|
|
Companies:
|
University of Michigan and University of Michigan
|
|
Keywords:
|
Graph-based dependency;
Multi-labeled variables ;
Computational complexity;
Feature selection ;
Biology application ;
Information theoretic measures
|
|
Abstract:
|
Several algorithms to learn the dependency between a pair of multivariate random variables directly from a data sample have been proposed in the past. This work proposes a new graph-based dependency criterion inspired by geometry of graphs and information theoretic measures to estimate dependency between multi-labels variables. The advantages of our proposed dependency estimator are demonstrated in a series of simulations. This approach results in an efficient and fast non-parametric implementation of dependency estimation with broad applications in biology. For instance, the proposed technique is applied to the genetic data set to filter out redundant features.
|
Authors who are presenting talks have a * after their name.
