Explaining Classification by Finding Response-Related Subgroups in Data

A method for explaining results of a regression based classifier is proposed. The data is clustered using a metric extracted from the classifier. This way, clusters found are related to classifier predictions, and each cluster can be considered a possible explanation for classification result. The c...

Full description

Saved in:
Bibliographic Details
Main Authors: Parviainen, Elina, Vehtari, Aki
Format: Conference Proceeding
Language:English
Subjects:
Biomedical computing
Biomedical engineering
Clustering algorithms
Computer networks
Concurrent computing
data abstraction
Data mining
Distributed computing
Input variables
MLP classifier
Predictive models
Software engineering
subgroup rules
supervised clustering
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:A method for explaining results of a regression based classifier is proposed. The data is clustered using a metric extracted from the classifier. This way, clusters found are related to classifier predictions, and each cluster can be considered a possible explanation for classification result. The clusters are described by simple rules, meant to be easy for a human to understand. The key points of the work are presenting a modular framework for explaining the classification, and studying and comparing two different approaches for extracting a metric from a classifier model.
DOI:10.1109/SNPD.2010.20