Deakin University
Browse

File(s) under permanent embargo

A study on reliability in graph discovery

conference contribution
posted on 2006-01-01, 00:00 authored by Honghua Dai
A critical question in data mining is that can we always trust what discovered by a data mining system unconditionally? The answer is obviously not. If not, when can we trust the discovery then? What are the factors that affect the reliability of the discovery? How do they affect the reliability of the discovery? These are some interesting questions to be investigated.

In this paper we will firstly provide a definition and the measurements of reliability, and analyse the factors that affect the reliability. We then examine the impact of model complexity, weak links, varying sample sizes and the ability of different learners to the reliability of graphical model discovery. The experimental results reveal that (1) the larger sample size for the discovery, the higher reliability we will get; (2) the stronger a graph link is, the easier the discovery will be and thus the higher the reliability it can achieve; (3) the complexity of a graph also plays an important role in the discovery. The higher the complexity of a graph is, the more difficult to induce the graph and the lower reliability it would be.

History

Event

International Conference on Data Mining ICDM (6th : 2006 : Hong Kong)

Pagination

775 - 779

Publisher

IEEE Computer Society

Location

Hong Kong

Place of publication

Los Alamitos, Calif.

Start date

2006-12-18

End date

2006-12-22

ISBN-13

9780769527024

ISBN-10

0769527027

Language

eng

Publication classification

E1 Full written paper - refereed

Copyright notice

2006, IEEE

Editor/Contributor(s)

C Clifton

Title of proceedings

Sixth International Conference on Data Mining ICDM 2006 : proceedings : 18-22 December, 2006, Hong Kong

Usage metrics

    Research Publications

    Categories

    No categories selected

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC