High accuracy partially monotone ordinal classication

Christopher Bartley

Research output: ThesisDoctoral Thesis

8 Downloads (Pure)

Abstract

This thesis aimed to answer the question: can domain knowledge regarding nondecreasing (monotone) relationships be incorporated into high performance classification algorithms without compromising their performance? Two high accuracy classifiers (Support Vector Machines and Random Forest) and a foundational approach (instance based partial orders) were extended into novel versions that demonstrated improvements in one or more of the following: global monotonicity (guaranteeing compliance in all circumstances), partial monotonicity (allowing non-monotone features), prediction accuracy, and scalability. It is hoped this thesis encourages and enables practitioners to include knowledge of monotonicity in their models, for the sake of accuracy, simplicity, and interpretability.
Original languageEnglish
QualificationDoctor of Philosophy
Awarding Institution
  • The University of Western Australia
Thesis sponsors
Award date13 Jun 2019
DOIs
Publication statusUnpublished - 2018

Fingerprint

Support vector machines
Scalability
Classifiers
Compliance

Cite this

@phdthesis{27a5fd56452d40f5bec22a58ab5638c6,
title = "High accuracy partially monotone ordinal classication",
abstract = "This thesis aimed to answer the question: can domain knowledge regarding nondecreasing (monotone) relationships be incorporated into high performance classification algorithms without compromising their performance? Two high accuracy classifiers (Support Vector Machines and Random Forest) and a foundational approach (instance based partial orders) were extended into novel versions that demonstrated improvements in one or more of the following: global monotonicity (guaranteeing compliance in all circumstances), partial monotonicity (allowing non-monotone features), prediction accuracy, and scalability. It is hoped this thesis encourages and enables practitioners to include knowledge of monotonicity in their models, for the sake of accuracy, simplicity, and interpretability.",
keywords = "Machine learning, Monotonicity, Domain knowledge, Classification",
author = "Christopher Bartley",
year = "2018",
doi = "10.26182/5d19aff34917e",
language = "English",
school = "The University of Western Australia",

}

Bartley, C 2018, 'High accuracy partially monotone ordinal classication', Doctor of Philosophy, The University of Western Australia. https://doi.org/10.26182/5d19aff34917e

High accuracy partially monotone ordinal classication. / Bartley, Christopher.

2018.

Research output: ThesisDoctoral Thesis

TY - THES

T1 - High accuracy partially monotone ordinal classication

AU - Bartley, Christopher

PY - 2018

Y1 - 2018

N2 - This thesis aimed to answer the question: can domain knowledge regarding nondecreasing (monotone) relationships be incorporated into high performance classification algorithms without compromising their performance? Two high accuracy classifiers (Support Vector Machines and Random Forest) and a foundational approach (instance based partial orders) were extended into novel versions that demonstrated improvements in one or more of the following: global monotonicity (guaranteeing compliance in all circumstances), partial monotonicity (allowing non-monotone features), prediction accuracy, and scalability. It is hoped this thesis encourages and enables practitioners to include knowledge of monotonicity in their models, for the sake of accuracy, simplicity, and interpretability.

AB - This thesis aimed to answer the question: can domain knowledge regarding nondecreasing (monotone) relationships be incorporated into high performance classification algorithms without compromising their performance? Two high accuracy classifiers (Support Vector Machines and Random Forest) and a foundational approach (instance based partial orders) were extended into novel versions that demonstrated improvements in one or more of the following: global monotonicity (guaranteeing compliance in all circumstances), partial monotonicity (allowing non-monotone features), prediction accuracy, and scalability. It is hoped this thesis encourages and enables practitioners to include knowledge of monotonicity in their models, for the sake of accuracy, simplicity, and interpretability.

KW - Machine learning

KW - Monotonicity

KW - Domain knowledge

KW - Classification

U2 - 10.26182/5d19aff34917e

DO - 10.26182/5d19aff34917e

M3 - Doctoral Thesis

ER -