One-class Machine Learning Approach for fMRI Analysis
One-class Machine Learning Approach for fMRI Analysis
One-Class Machine Learning techniques (i.e. "bottleneck" neural networks and one-class support vector machines (SVM)) are applied to classify whether a subject is performing a task or not by looking solely at the raw fMRI slices of his brain. "One-class" means that during training the system only has access to positive (i.e. task performing) examples. "Two-class" means it has access to negative examples as well. Successful classification of data by a system trained under either of the one-class systems was accomplished at close to the 60% level. (In contrast, an implementation of a standard two class SVM succeeds at around the 70% level.) These results were stable over repeated experiments and for both motor and visual tasks. Since the one-class neural network technique is naturally related to dimension reduction, it is possible that this mechanism may also be used for feature selection.
fMRI, SVM, Neural Networks, Machine Learning, One-Class
Hardoon, David R
ba75cf10-43f2-4701-b648-3bb90edbabbf
Manevitz, Larry M
56d193a6-d111-43db-a5d2-c56ee728dabf
2005
Hardoon, David R
ba75cf10-43f2-4701-b648-3bb90edbabbf
Manevitz, Larry M
56d193a6-d111-43db-a5d2-c56ee728dabf
Hardoon, David R and Manevitz, Larry M
(2005)
One-class Machine Learning Approach for fMRI Analysis.
Postgraduate Research Conference in Electronics, Photonics, Communications and Networks, and Computer Science, Lancaster, United Kingdom.
30 Mar - 01 Apr 2005.
Record type:
Conference or Workshop Item
(Poster)
Abstract
One-Class Machine Learning techniques (i.e. "bottleneck" neural networks and one-class support vector machines (SVM)) are applied to classify whether a subject is performing a task or not by looking solely at the raw fMRI slices of his brain. "One-class" means that during training the system only has access to positive (i.e. task performing) examples. "Two-class" means it has access to negative examples as well. Successful classification of data by a system trained under either of the one-class systems was accomplished at close to the 60% level. (In contrast, an implementation of a standard two class SVM succeeds at around the 70% level.) These results were stable over repeated experiments and for both motor and visual tasks. Since the one-class neural network technique is naturally related to dimension reduction, it is possible that this mechanism may also be used for feature selection.
Text
Hardoon_PREP05.pdf
- Other
More information
Published date: 2005
Additional Information:
Event Dates: 30 March - 01 April
Venue - Dates:
Postgraduate Research Conference in Electronics, Photonics, Communications and Networks, and Computer Science, Lancaster, United Kingdom, 2005-03-30 - 2005-04-01
Keywords:
fMRI, SVM, Neural Networks, Machine Learning, One-Class
Organisations:
Electronics & Computer Science
Identifiers
Local EPrints ID: 260661
URI: http://eprints.soton.ac.uk/id/eprint/260661
PURE UUID: 80666e09-8c43-4dfb-8b04-e38d98f10b90
Catalogue record
Date deposited: 08 Mar 2005
Last modified: 14 Mar 2024 06:41
Export record
Contributors
Author:
David R Hardoon
Author:
Larry M Manevitz
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics