Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
6742.pdf (609.75 KB)
ETD Abstract Container
Abstract Header
Methodology For Generating High-Confidence Cost-Sensitive Rules For Classification
Author Info
Bakshi, Arjun
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1377868085
Abstract Details
Year and Degree
2013, MS, University of Cincinnati, Engineering and Applied Science: Computer Science.
Abstract
Rule based classifiers are often used to make crucial decision in domains like medicine and business intelligence, where there is a need to build insightful models that are quick to train, perform accurate classification, and take the costs of mistakes into account while making or helping with predictions. The existing techniques that address these requirements suffer from some disadvantages that cause them to generate overly complicated rule sets that sometimes do not perform well on new data, or do not take differing misclassification costs into account. The work proposed here aims to build a rule based classifier that extracts rules that have higher support and confidence than existing techniques as well as a classification model that minimizes the cost incurred from misclassifications by making cost sensitive decisions and flagging instances that are likely to be misclassified.
Committee
Raj Bhatnagar, Ph.D. (Committee Chair)
Anil Jegga, D.V.M., M.Res. (Committee Member)
Ali Minai, Ph.D. (Committee Member)
Pages
67 p.
Subject Headings
Computer Science
Keywords
Cost sensitive
;
High Confidence
;
Association Rules
;
Classification
;
Non forced
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Bakshi, A. (2013).
Methodology For Generating High-Confidence Cost-Sensitive Rules For Classification
[Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1377868085
APA Style (7th edition)
Bakshi, Arjun.
Methodology For Generating High-Confidence Cost-Sensitive Rules For Classification.
2013. University of Cincinnati, Master's thesis.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1377868085.
MLA Style (8th edition)
Bakshi, Arjun. "Methodology For Generating High-Confidence Cost-Sensitive Rules For Classification." Master's thesis, University of Cincinnati, 2013. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1377868085
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
ucin1377868085
Download Count:
421
Copyright Info
© 2013, some rights reserved.
Methodology For Generating High-Confidence Cost-Sensitive Rules For Classification by Arjun Bakshi is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License. Based on a work at etd.ohiolink.edu.
This open access ETD is published by University of Cincinnati and OhioLINK.