UNSUPERVISED LEARNING APPROACH TO FEATURE SELECTION IN BIOLOGICAL DATA ANALYSIS

Witold Jacak; Karin Pröll

UNSUPERVISED LEARNING APPROACH TO FEATURE SELECTION IN BIOLOGICAL DATA ANALYSIS

Witold Jacak, Karin Pröll

Research Center Hagenberg

Research output: Chapter in Book/Report/Conference proceedings › Conference contribution

Abstract

In this paper we present a novel method for scoring function specification and feature selection by combining unsupervised learning with supervised cross validation. Unsupervised clustering methods (k-means, one dimensional Kohonen SOM, fuzzy c-means) are used to perform a clustering of object-data for a chosen subset of input features and given number of clusters. The resulting object clusters are compared with the predefined original object classes and a matching factor (score) is calculated. This score is used as criterion function for heuristic sequential feature selection and novel cross selection algorithm.

Original language	English
Title of host publication	24th European Modeling and Simulation Symposium, EMSS 2012
Pages	232-236
Number of pages	5
Publication status	Published - 2012
Event	The 24th European Modeling & Simulation Symposium (EMSS 2012) - Vienna, Austria Duration: 19 Sept 2012 → 21 Sept 2012 http://www.msc-les.org/conf/EMSS2012/

Publication series

Name	24th European Modeling and Simulation Symposium, EMSS 2012

Conference

Conference	The 24th European Modeling & Simulation Symposium (EMSS 2012)
Country/Territory	Austria
City	Vienna
Period	19.09.2012 → 21.09.2012
Internet address	http://www.msc-les.org/conf/EMSS2012/

Keywords

classification
clustering
feature selection
k-mean
fuzzy c-mean
Kohonen SOM
MLP
Feature selection
Classification
Clustering
Fuzzy c-mean
K-mean

Cite this

@inproceedings{07c210f9d1be44fa8d3c37edfa4cc204,

title = "UNSUPERVISED LEARNING APPROACH TO FEATURE SELECTION IN BIOLOGICAL DATA ANALYSIS",

abstract = "In this paper we present a novel method for scoring function specification and feature selection by combining unsupervised learning with supervised cross validation. Unsupervised clustering methods (k-means, one dimensional Kohonen SOM, fuzzy c-means) are used to perform a clustering of object-data for a chosen subset of input features and given number of clusters. The resulting object clusters are compared with the predefined original object classes and a matching factor (score) is calculated. This score is used as criterion function for heuristic sequential feature selection and novel cross selection algorithm.",

keywords = "classification, clustering, feature selection, k-mean, fuzzy c-mean, Kohonen SOM, MLP, classification, clustering, feature selection, k-mean, fuzzy c-mean, Kohonen SOM, MLP, Feature selection, Classification, Clustering, Fuzzy c-mean, K-mean",

author = "Witold Jacak and Karin Pr{\"o}ll",

year = "2012",

language = "English",

isbn = "9788897999096",

series = "24th European Modeling and Simulation Symposium, EMSS 2012",

pages = "232--236",

booktitle = "24th European Modeling and Simulation Symposium, EMSS 2012",

note = "The 24th European Modeling & Simulation Symposium (EMSS 2012) ; Conference date: 19-09-2012 Through 21-09-2012",

url = "http://www.msc-les.org/conf/EMSS2012/",

}

TY - GEN

T1 - UNSUPERVISED LEARNING APPROACH TO FEATURE SELECTION IN BIOLOGICAL DATA ANALYSIS

AU - Jacak, Witold

AU - Pröll, Karin

PY - 2012

Y1 - 2012

N2 - In this paper we present a novel method for scoring function specification and feature selection by combining unsupervised learning with supervised cross validation. Unsupervised clustering methods (k-means, one dimensional Kohonen SOM, fuzzy c-means) are used to perform a clustering of object-data for a chosen subset of input features and given number of clusters. The resulting object clusters are compared with the predefined original object classes and a matching factor (score) is calculated. This score is used as criterion function for heuristic sequential feature selection and novel cross selection algorithm.

AB - In this paper we present a novel method for scoring function specification and feature selection by combining unsupervised learning with supervised cross validation. Unsupervised clustering methods (k-means, one dimensional Kohonen SOM, fuzzy c-means) are used to perform a clustering of object-data for a chosen subset of input features and given number of clusters. The resulting object clusters are compared with the predefined original object classes and a matching factor (score) is calculated. This score is used as criterion function for heuristic sequential feature selection and novel cross selection algorithm.

KW - classification

KW - clustering

KW - feature selection

KW - k-mean

KW - fuzzy c-mean

KW - Kohonen SOM

KW - MLP

KW - classification

KW - clustering

KW - feature selection

KW - k-mean

KW - fuzzy c-mean

KW - Kohonen SOM

KW - MLP

KW - Feature selection

KW - Classification

KW - Clustering

KW - Fuzzy c-mean

KW - K-mean

UR - http://www.scopus.com/inward/record.url?scp=84871515022&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9788897999096

T3 - 24th European Modeling and Simulation Symposium, EMSS 2012

SP - 232

EP - 236

BT - 24th European Modeling and Simulation Symposium, EMSS 2012

T2 - The 24th European Modeling & Simulation Symposium (EMSS 2012)

Y2 - 19 September 2012 through 21 September 2012

ER -

UNSUPERVISED LEARNING APPROACH TO FEATURE SELECTION IN BIOLOGICAL DATA ANALYSIS

Abstract

Publication series

Conference

Keywords

Other files and links

Fingerprint

Cite this