AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble

Jing-Jing CAO; Sam KWONG; Ran WANG; Ke LI

doi:10.1109/ICMLC.2014.7009108

AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble

Jing-Jing CAO, Sam KWONG, Ran WANG, Ke LI

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

2 Citations (Scopus)

Abstract

One of the most difficult components for multi-class classification system is to find an appropriate error-correcting output codes (ECOC) matrix, which is used to decompose the multi-class problem into several binary class problems. In this paper, an indicator based multi-objective evolutionary algorithm with preference involved is designed to search the high-quality ECOC matrix. Specifically, the Harrington's one-sided desirability function is integrated into an indicator-based evolutionary algorithm (IBEA), which aims to approximate the relevant regions of pareto front (PF) according to the preference of the decision maker. Simulation results show that the proposed approach has better classification performance than compared multi-class based algorithms

Original language	English
Title of host publication	Proceedings - International Conference on Machine Learning and Cybernetics
Pages	147-152
DOIs	https://doi.org/10.1109/ICMLC.2014.7009108
Publication status	Published - 13 Jan 2015
Externally published	Yes

Keywords

Error-correcting output coding
Harrington's one-sided desirability function
Indicator-based evolutionary algorithm
Multi-class problem
Pareto front

Access to Document

10.1109/ICMLC.2014.7009108

Cite this

@inproceedings{f1dd04079bbc40d8981637f5db89663f,

title = "AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble",

abstract = "One of the most difficult components for multi-class classification system is to find an appropriate error-correcting output codes (ECOC) matrix, which is used to decompose the multi-class problem into several binary class problems. In this paper, an indicator based multi-objective evolutionary algorithm with preference involved is designed to search the high-quality ECOC matrix. Specifically, the Harrington's one-sided desirability function is integrated into an indicator-based evolutionary algorithm (IBEA), which aims to approximate the relevant regions of pareto front (PF) according to the preference of the decision maker. Simulation results show that the proposed approach has better classification performance than compared multi-class based algorithms",

keywords = "Error-correcting output coding, Harrington's one-sided desirability function, Indicator-based evolutionary algorithm, Multi-class problem, Pareto front",

author = "Jing-Jing CAO and Sam KWONG and Ran WANG and Ke LI",

year = "2015",

month = jan,

day = "13",

doi = "10.1109/ICMLC.2014.7009108",

language = "English",

pages = "147--152",

booktitle = "Proceedings - International Conference on Machine Learning and Cybernetics",

}

AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble. / CAO, Jing-Jing; KWONG, Sam; WANG, Ran et al.
Proceedings - International Conference on Machine Learning and Cybernetics. 2015. p. 147-152.

Research output: Book Chapters | Papers in Conference Proceedings › Conference paper (refereed) › Research › peer-review

TY - GEN

T1 - AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble

AU - CAO, Jing-Jing

AU - KWONG, Sam

AU - WANG, Ran

AU - LI, Ke

PY - 2015/1/13

Y1 - 2015/1/13

N2 - One of the most difficult components for multi-class classification system is to find an appropriate error-correcting output codes (ECOC) matrix, which is used to decompose the multi-class problem into several binary class problems. In this paper, an indicator based multi-objective evolutionary algorithm with preference involved is designed to search the high-quality ECOC matrix. Specifically, the Harrington's one-sided desirability function is integrated into an indicator-based evolutionary algorithm (IBEA), which aims to approximate the relevant regions of pareto front (PF) according to the preference of the decision maker. Simulation results show that the proposed approach has better classification performance than compared multi-class based algorithms

AB - One of the most difficult components for multi-class classification system is to find an appropriate error-correcting output codes (ECOC) matrix, which is used to decompose the multi-class problem into several binary class problems. In this paper, an indicator based multi-objective evolutionary algorithm with preference involved is designed to search the high-quality ECOC matrix. Specifically, the Harrington's one-sided desirability function is integrated into an indicator-based evolutionary algorithm (IBEA), which aims to approximate the relevant regions of pareto front (PF) according to the preference of the decision maker. Simulation results show that the proposed approach has better classification performance than compared multi-class based algorithms

KW - Error-correcting output coding

KW - Harrington's one-sided desirability function

KW - Indicator-based evolutionary algorithm

KW - Multi-class problem

KW - Pareto front

UR - http://www.scopus.com/inward/record.url?scp=84921506936&partnerID=8YFLogxK

U2 - 10.1109/ICMLC.2014.7009108

DO - 10.1109/ICMLC.2014.7009108

M3 - Conference paper (refereed)

SP - 147

EP - 152

BT - Proceedings - International Conference on Machine Learning and Cybernetics

ER -

AN indicator-based selection multi-objective evolutionary algorithm with preference for multi-class ensemble

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this