ATTENTION: The software behind KU ScholarWorks is being upgraded to a new version. Starting July 15th, users will not be able to log in to the system, add items, nor make any changes until the new version is in place at the end of July. Searching for articles and opening files will continue to work while the system is being updated. If you have any questions, please contact Marianne Reed at mreed@ku.edu .

Show simple item record

dc.contributor.advisorGrzymala-Busse, Jerzy W.
dc.contributor.authorChoudhury Arunabha
dc.date.accessioned2015-09-07T22:29:22Z
dc.date.available2015-09-07T22:29:22Z
dc.date.issued2014-12-31
dc.date.submitted2014
dc.identifier.otherhttp://dissertations.umi.com/ku:13736
dc.identifier.urihttp://hdl.handle.net/1808/18418
dc.description.abstractThis work formally introduces a generalized fuzzy logic and interval clustering (FLIC) technique which, when integrated with existing supervised learning algorithms, improves their performance. FLIC is a method that was first integrated with neural network in order to improve neural network's performance in drug discovery using high throughput screening (HTS). This research strictly focuses on binary classification problems and generalizes the FLIC in order to incorporate it with other machine learning algorithms. In most binary classification problems, the class boundary is not linear. This pose a major problem when the number of outliers are significantly high, degrading the performance of the supervised learning function. FLIC identifies these misclassifications before the training set is introduced to the learning algorithm. This allows the supervised learning algorithm to learn more efficiently since it is now aware of those misclassifications. Although the proposed method performs well with most binary classification problems, it does significantly well for data set with high class asymmetry. The proposed method has been tested on four well known data sets of which three are from UCI Machine Learning repository and one from BigML. Tests have been conducted with three well known supervised learning techniques: Decision Tree, Logistic Regression and Naive Bayes. The results from the experiments show significant improvement in performance. The paper begins with a formal introduction to the core idea this research is based upon. It then discusses a list of other methods that have either inspired this research or have been referred to, in order to formalize the techniques. Subsequent sections discuss the methodology and the algorithm which is followed by results and conclusion.
dc.format.extent71 pages
dc.language.isoen
dc.publisherUniversity of Kansas
dc.rightsCopyright held by the author.
dc.subjectComputer science
dc.subjectBinary Classifiers
dc.subjectClustering
dc.subjectData Mining
dc.subjectFuzzy Logic
dc.subjectMachine Learning
dc.subjectSupervised Learning
dc.titleGeneralized FLIC: Learning with Misclassification for Binary Classifiers
dc.typeThesis
dc.contributor.cmtememberChakrabarti, Swapan
dc.contributor.cmtememberLuo, Bo
dc.thesis.degreeDisciplineElectrical Engineering & Computer Science
dc.thesis.degreeLevelM.S.
dc.rights.accessrightsopenAccess


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record