Optimizing between data transformation and parametric weighting for stable binary classification

Kangrok Oh, Zhengguo Li, Beom Seok Oh, Kar Ann Toh

Research output: Contribution to journalArticle

Abstract

In this paper, an optimization problem is formulated for stable binary classification. Essentially, the objective function seeks to optimize a full data transformation matrix along with the learning of a linear parametric model. The data transformation matrix and the weight parameter vector are alternatingly optimized based on the area above the receiver operating characteristic curve criterion. The proposed method improves the existing means via an optimal data transformation rather than that based on the diagonal, random and ad-hoc settings. This optimal transformation stretches beyond the fixed settings of known optimization methods. Extensive experiments using 34 binary classification data sets show that the proposed method can be more stable than competing classifiers. Specifically, the proposed method shows robustness to imbalanced and small training data sizes in terms of classification accuracy with statistical evidence.

Original languageEnglish
Pages (from-to)1614-1637
Number of pages24
JournalJournal of the Franklin Institute
Volume355
Issue number4
DOIs
Publication statusPublished - 2018 Mar 1

Fingerprint

Data Transformation
Binary Classification
Weighting
Transformation Matrix
Receiver Operating Characteristic Curve
Stretch
Parametric Model
Optimization Methods
Linear Model
Classifiers
Objective function
Classifier
Optimise
Robustness
Optimization Problem
Experiment
Experiments

All Science Journal Classification (ASJC) codes

  • Control and Systems Engineering
  • Signal Processing
  • Computer Networks and Communications
  • Applied Mathematics

Cite this

@article{aa69341733f145519aae7367078e0568,
title = "Optimizing between data transformation and parametric weighting for stable binary classification",
abstract = "In this paper, an optimization problem is formulated for stable binary classification. Essentially, the objective function seeks to optimize a full data transformation matrix along with the learning of a linear parametric model. The data transformation matrix and the weight parameter vector are alternatingly optimized based on the area above the receiver operating characteristic curve criterion. The proposed method improves the existing means via an optimal data transformation rather than that based on the diagonal, random and ad-hoc settings. This optimal transformation stretches beyond the fixed settings of known optimization methods. Extensive experiments using 34 binary classification data sets show that the proposed method can be more stable than competing classifiers. Specifically, the proposed method shows robustness to imbalanced and small training data sizes in terms of classification accuracy with statistical evidence.",
author = "Kangrok Oh and Zhengguo Li and Oh, {Beom Seok} and Toh, {Kar Ann}",
year = "2018",
month = "3",
day = "1",
doi = "10.1016/j.jfranklin.2017.04.012",
language = "English",
volume = "355",
pages = "1614--1637",
journal = "Journal of the Franklin Institute",
issn = "0016-0032",
publisher = "Elsevier Limited",
number = "4",

}

Optimizing between data transformation and parametric weighting for stable binary classification. / Oh, Kangrok; Li, Zhengguo; Oh, Beom Seok; Toh, Kar Ann.

In: Journal of the Franklin Institute, Vol. 355, No. 4, 01.03.2018, p. 1614-1637.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Optimizing between data transformation and parametric weighting for stable binary classification

AU - Oh, Kangrok

AU - Li, Zhengguo

AU - Oh, Beom Seok

AU - Toh, Kar Ann

PY - 2018/3/1

Y1 - 2018/3/1

N2 - In this paper, an optimization problem is formulated for stable binary classification. Essentially, the objective function seeks to optimize a full data transformation matrix along with the learning of a linear parametric model. The data transformation matrix and the weight parameter vector are alternatingly optimized based on the area above the receiver operating characteristic curve criterion. The proposed method improves the existing means via an optimal data transformation rather than that based on the diagonal, random and ad-hoc settings. This optimal transformation stretches beyond the fixed settings of known optimization methods. Extensive experiments using 34 binary classification data sets show that the proposed method can be more stable than competing classifiers. Specifically, the proposed method shows robustness to imbalanced and small training data sizes in terms of classification accuracy with statistical evidence.

AB - In this paper, an optimization problem is formulated for stable binary classification. Essentially, the objective function seeks to optimize a full data transformation matrix along with the learning of a linear parametric model. The data transformation matrix and the weight parameter vector are alternatingly optimized based on the area above the receiver operating characteristic curve criterion. The proposed method improves the existing means via an optimal data transformation rather than that based on the diagonal, random and ad-hoc settings. This optimal transformation stretches beyond the fixed settings of known optimization methods. Extensive experiments using 34 binary classification data sets show that the proposed method can be more stable than competing classifiers. Specifically, the proposed method shows robustness to imbalanced and small training data sizes in terms of classification accuracy with statistical evidence.

UR - http://www.scopus.com/inward/record.url?scp=85018972073&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85018972073&partnerID=8YFLogxK

U2 - 10.1016/j.jfranklin.2017.04.012

DO - 10.1016/j.jfranklin.2017.04.012

M3 - Article

AN - SCOPUS:85018972073

VL - 355

SP - 1614

EP - 1637

JO - Journal of the Franklin Institute

JF - Journal of the Franklin Institute

SN - 0016-0032

IS - 4

ER -