Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis

Romesh Ranawana; Vasile Palade; Daniel Howard

doi:10.1109/FBIT.2007.146

Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis

Romesh Ranawana, Vasile Palade, Daniel Howard

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding › peer-review

4 Citations (Scopus)

Abstract

Learning algorithms aim for accuracy of classification but this depends on a choice of heuristic metric to measure performance and also on the proper consideration and addressing of the important requirements of the classification task. This paper introduces a framework, MVGen, to implement different training heuristics capable of inducing the training algorithm that can provide the desired results while negating detrimental aspects of a training set imbalance. Our experiments indicate that successful classifiers can indeed be built to specialize on the minority class within an imbalanced data set.

Original language	English
Title of host publication	Proceedings of the Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007
Publisher	IEEE
Pages	341-346
Number of pages	6
ISBN (Print)	0769529992, 9780769529998
DOIs	https://doi.org/10.1109/FBIT.2007.146
Publication status	Published - 2007
Externally published	Yes
Event	Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007 - Jeju Island, Korea, Republic of Duration: 11 Oct 2007 → 13 Oct 2007

Conference

Conference	Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007
Country/Territory	Korea, Republic of
City	Jeju Island
Period	11/10/07 → 13/10/07

Keywords

Genetic algorithms
Algorithm design and analysis
Testing
Sequences
DNA computing
Laboratories
Neural networks
Guidelines
Bioinformatics
Information technology
genetic algorithms
biology computing
DNA
imbalanced data set
genetic algorithm approach
specialized multiclassifier systems
DNA analysis
learning algorithms
training algorithm

ASJC Scopus subject areas

Computer Science Applications
Information Systems

Access to Document

10.1109/FBIT.2007.146

Cite this

Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis. / Ranawana, Romesh; Palade, Vasile; Howard, Daniel.
Proceedings of the Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007. IEEE, 2007. p. 341-346 4524130.

Research output: Chapter in Book/Report/Conference proceeding › Conference proceeding › peer-review

Ranawana, R, Palade, V & Howard, D 2007, Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis. in Proceedings of the Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007., 4524130, IEEE, pp. 341-346, Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007, Jeju Island, Korea, Republic of, 11/10/07. https://doi.org/10.1109/FBIT.2007.146

@inproceedings{5b4e64a684aa43699f80dbde56f1183c,

title = "Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis",

abstract = "Learning algorithms aim for accuracy of classification but this depends on a choice of heuristic metric to measure performance and also on the proper consideration and addressing of the important requirements of the classification task. This paper introduces a framework, MVGen, to implement different training heuristics capable of inducing the training algorithm that can provide the desired results while negating detrimental aspects of a training set imbalance. Our experiments indicate that successful classifiers can indeed be built to specialize on the minority class within an imbalanced data set.",

keywords = "Genetic algorithms, Algorithm design and analysis, Testing, Sequences, DNA computing, Laboratories, Neural networks, Guidelines, Bioinformatics, Information technology, genetic algorithms, biology computing, DNA, imbalanced data set, genetic algorithm approach, specialized multiclassifier systems, DNA analysis, learning algorithms, training algorithm",

author = "Romesh Ranawana and Vasile Palade and Daniel Howard",

year = "2007",

doi = "10.1109/FBIT.2007.146",

language = "English",

isbn = "0769529992",

pages = "341--346",

booktitle = "Proceedings of the Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007",

publisher = "IEEE",

address = "United States",

note = "Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007 ; Conference date: 11-10-2007 Through 13-10-2007",

}

TY - GEN

T1 - Genetic algorithm approach to construction of specialized multi-classifier systems

T2 - Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007

AU - Ranawana, Romesh

AU - Palade, Vasile

AU - Howard, Daniel

PY - 2007

Y1 - 2007

N2 - Learning algorithms aim for accuracy of classification but this depends on a choice of heuristic metric to measure performance and also on the proper consideration and addressing of the important requirements of the classification task. This paper introduces a framework, MVGen, to implement different training heuristics capable of inducing the training algorithm that can provide the desired results while negating detrimental aspects of a training set imbalance. Our experiments indicate that successful classifiers can indeed be built to specialize on the minority class within an imbalanced data set.

AB - Learning algorithms aim for accuracy of classification but this depends on a choice of heuristic metric to measure performance and also on the proper consideration and addressing of the important requirements of the classification task. This paper introduces a framework, MVGen, to implement different training heuristics capable of inducing the training algorithm that can provide the desired results while negating detrimental aspects of a training set imbalance. Our experiments indicate that successful classifiers can indeed be built to specialize on the minority class within an imbalanced data set.

KW - Genetic algorithms

KW - Algorithm design and analysis

KW - Testing

KW - Sequences

KW - DNA computing

KW - Laboratories

KW - Neural networks

KW - Guidelines

KW - Bioinformatics

KW - Information technology

KW - genetic algorithms

KW - biology computing

KW - DNA

KW - imbalanced data set

KW - genetic algorithm approach

KW - specialized multiclassifier systems

KW - DNA analysis

KW - learning algorithms

KW - training algorithm

UR - http://www.scopus.com/inward/record.url?scp=49349099642&partnerID=8YFLogxK

U2 - 10.1109/FBIT.2007.146

DO - 10.1109/FBIT.2007.146

M3 - Conference proceeding

AN - SCOPUS:49349099642

SN - 0769529992

SN - 9780769529998

SP - 341

EP - 346

BT - Proceedings of the Frontiers in the Convergence of Bioscience and Information Technologies, FBIT 2007

PB - IEEE

Y2 - 11 October 2007 through 13 October 2007

ER -

Genetic algorithm approach to construction of specialized multi-classifier systems: Application to DNA analysis

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this