Files in this item

FilesDescriptionFormat

application/pdf

application/pdfALEBIOSU-THESIS-2017.pdf (452kB)
(no description provided)PDF

Description

Title:Multiple-implementation testing of supervised learning software
Author(s):Alebiosu, Oreoluwa
Advisor(s):Xie, Tao
Department / Program:Computer Science
Discipline:Computer Science
Degree Granting Institution:University of Illinois at Urbana-Champaign
Degree:M.S.
Genre:Thesis
Subject(s):Machine learning
multiple-implementation testing
differential testing
supervised learning
multiple-implementation monitoring
software monitoring
k-nearest neighbor
knn
naive bayes
naivebayes
software testing
pseudo oracle
algorithm configurations
percentage threshold
black box
test oracle
multiple implementation
Abstract:Machine Learning (ML) software, used to implement an ML algorithm, is widely used in many application domains such as financial, business, and engineering domains. Faults in ML software can cause substantial losses in these application domains. Thus, it is very critical to conduct effective testing of ML software to detect and eliminate its faults. However, testing ML software is difficult, especially on producing test oracles used for checking behavior correctness (such as using expected properties or expected test outputs). To tackle the test-oracle issue, this thesis presents a novel black-box approach of multiple-implementation testing for supervised learning software. The insight underlying the approach is that there can be multiple implementations (independently written) for a supervised learning algorithm, and majority of them may produce the expected output for a test input (even if none of these implementations are fault-free). In particular, the proposed approach derives a pseudo oracle for a test input by running the test input on n implementations of the supervised learning algorithm, and then using the common test output produced by a majority (determined by a percentage threshold) of these n implementations. The proposed approach includes techniques to address challenges in multiple-implementation testing (or generally testing) of supervised learning software: the definition of test cases in testing supervised learning software, along with resolution of inconsistent algorithm configurations across implementations. In addition, to improve dependability of supervised learning software during in-field usage while incurring low runtime overhead, The approach includes a multiple-implementation monitoring technique. The evaluations on the proposed approach show that multiple-implementation testing is effective in detecting real faults in real-world ML software (even popularly used ones), including 5 faults from 10 NaiveBayes implementations and 4 faults from 20 k-nearest neighbor implementations, and the proposed technique of multiple-implementation monitoring substantially reduces the need of running multiple implementations with high prediction accuracy.
Issue Date:2017-04-26
Type:Thesis
URI:http://hdl.handle.net/2142/97479
Rights Information:Copyright 2017 Oreoluwa Alebiosu
Date Available in IDEALS:2017-08-10
Date Deposited:2017-05


This item appears in the following Collection(s)

Item Statistics