Random Subspace Ensemble Artificial Neural Networks for
Firstepisode Schizophrenia Classification

D 2016

Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification

VYŠKOVSKÝ, Roman, Daniel SCHWARZ, Eva JANOUŠOVÁ and Tomáš KAŠPÁREK

Basic information

Original name

Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification

Authors

VYŠKOVSKÝ, Roman (203 Czech Republic, belonging to the institution), Daniel SCHWARZ (203 Czech Republic, belonging to the institution), Eva JANOUŠOVÁ (203 Czech Republic, belonging to the institution) and Tomáš KAŠPÁREK (203 Czech Republic, belonging to the institution)

Edition

Warzaw; Los Alamitos, Annals of Computer Science and Information Systems, Volume 8 : Proceedings of the 2016 Federated Conference on Computer Science and Information Systems, p. 317-321, 5 pp. 2016

Publisher

Polskie Towarzystwo Informatyczne; Institute of Electrical and Electronics Engineers

Other information

Language

English

Type of outcome

Stať ve sborníku

Field of Study

10201 Computer sciences, information science, bioinformatics

Country of publisher

Poland

Confidentiality degree

není předmětem státního či obchodního tajemství

Publication form

printed version "print"

References:

URL

RIV identification code

RIV/00216224:14110/16:00091907

Organization unit

Faculty of Medicine

ISBN

978-83-60810-90-3

ISSN

DOI

http://dx.doi.org/10.15439/2016F333

UT WoS

000392436600050

Keywords in English

first-episode schizophrenia

Abstract

V originále

Computer-aided schizophrenia diagnosis is a difficult task that has been developing for last decades. Since traditional classifiers have not reached sufficient sensitivity and specificity, another possible way is combining the classifiers in ensembles. In this paper, we take advantage of random subspace ensemble method and combine it with multilayer perceptron (MLP) and support vector machines (SVM). Our experiment employs voxel-based morphometry to extract the grey matter densities from 52 images of first-episode schizophrenia patients and 52 healthy controls. MLP and SVM are adapted on random feature vectors taken from predefined feature pool and the classification results are based on their voting. Random feature ensemble method improved prediction of schizophrenia when short input feature vector (100 features) was used, however the performance was comparable with single classifiers based on bigger input feature vector (1000 and 10000 features).

Citovat

VYŠKOVSKÝ, Roman, Daniel SCHWARZ, Eva JANOUŠOVÁ and Tomáš KAŠPÁREK. Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification. In Maria Ganzha, Leszek Maciaszek, Marcin Paprzycki. Annals of Computer Science and Information Systems, Volume 8 : Proceedings of the 2016 Federated Conference on Computer Science and Information Systems. Warzaw; Los Alamitos: Polskie Towarzystwo Informatyczne; Institute of Electrical and Electronics Engineers, 2016, p. 317-321. ISBN 978-83-60810-90-3. Available from: https://dx.doi.org/10.15439/2016F333.

@inproceedings{1361770,
   author = {Vyškovský, Roman and Schwarz, Daniel and Janoušová, Eva and Kašpárek, Tomáš},
   address = {Warzaw; Los Alamitos},
   booktitle = {Annals of Computer Science and Information Systems, Volume 8 : Proceedings of the 2016 Federated Conference on Computer Science and Information Systems},
   doi = {http://dx.doi.org/10.15439/2016F333},
   editor = {Maria Ganzha, Leszek Maciaszek, Marcin Paprzycki},
   keywords = {first-episode schizophrenia},
   howpublished = {tištěná verze "print"},
   language = {eng},
   location = {Warzaw; Los Alamitos},
   isbn = {978-83-60810-90-3},
   pages = {317-321},
   publisher = {Polskie Towarzystwo Informatyczne; Institute of Electrical and Electronics Engineers},
   title = {Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification},
   url = {https://fedcsis.org/proceedings/2016/drp/333.html},
   year = {2016}
}

TY  - JOUR
ID  - 1361770
AU  - Vyškovský, Roman - Schwarz, Daniel - Janoušová, Eva - Kašpárek, Tomáš
PY  - 2016
TI  - Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification
PB  - Polskie Towarzystwo Informatyczne; Institute of Electrical and Electronics Engineers
CY  - Warzaw; Los Alamitos
SN  - 9788360810903
KW  - first-episode schizophrenia
UR  - https://fedcsis.org/proceedings/2016/drp/333.html
L2  - https://fedcsis.org/proceedings/2016/drp/333.html
N2  - Computer-aided schizophrenia diagnosis is a difficult task that has been developing for last decades. Since traditional classifiers have not reached sufficient sensitivity and specificity, another possible way is combining the classifiers in ensembles. In this paper, we take advantage of random subspace ensemble method and combine it with multilayer perceptron (MLP) and support vector machines (SVM). Our experiment employs voxel-based morphometry to extract the grey matter densities from 52 images of first-episode schizophrenia patients and 52 healthy controls. MLP and SVM are adapted on random feature vectors taken from predefined feature pool and the classification results are based on their voting. Random feature ensemble method improved prediction of schizophrenia when short input feature vector (100 features) was used, however the performance was comparable with single classifiers based on bigger input feature vector (1000 and 10000 features).
ER  -

VYŠKOVSKÝ, Roman, Daniel SCHWARZ, Eva JANOUŠOVÁ and Tomáš KAŠPÁREK. Random Subspace Ensemble Artificial Neural Networks for Firstepisode Schizophrenia Classification. In Maria Ganzha, Leszek Maciaszek, Marcin Paprzycki. \textit{Annals of Computer Science and Information Systems, Volume 8 : Proceedings of the 2016 Federated Conference on Computer Science and Information Systems}. Warzaw; Los Alamitos: Polskie Towarzystwo Informatyczne; Institute of Electrical and Electronics Engineers, 2016, p.~317-321. ISBN~978-83-60810-90-3. Available from: https://dx.doi.org/10.15439/2016F333.

Detailed Information on Publication Record