Abstract
Building good feature representations and learning hidden source models typically requires large sample sizes. In many applications, however, the size of the sample at an individual data holder may not be sufficient. One such application is neuroimaging analyses for mental health disorders - there are many individual research groups, each with a moderate number of subjects. Pooling such data can enable efficient feature learning, but privacy concerns prevent sharing the underlying data. We propose a model for private feature learning in which the data holders share differentially private views of their respective datasets to enable collaborative learning of a joint feature map. We give an example of such an algorithm for independent component analysis (ICA) - a popular blind source separation algorithm used in neuroimaging analyses. Our algorithm is a differentially private version of the recently proposed distributed joint ICA algorithm. We evaluate the performance of this method on simulated functional magnetic resonance imaging (fMRI) data.
| Original language | English (US) |
|---|---|
| Title of host publication | 2016 50th Annual Conference on Information Systems and Sciences, CISS 2016 |
| Publisher | Institute of Electrical and Electronics Engineers Inc. |
| Pages | 123-127 |
| Number of pages | 5 |
| ISBN (Electronic) | 9781467394574 |
| DOIs | |
| State | Published - Apr 26 2016 |
| Externally published | Yes |
| Event | 50th Annual Conference on Information Systems and Sciences, CISS 2016 - Princeton, United States Duration: Mar 16 2016 → Mar 18 2016 |
Other
| Other | 50th Annual Conference on Information Systems and Sciences, CISS 2016 |
|---|---|
| Country | United States |
| City | Princeton |
| Period | 3/16/16 → 3/18/16 |
Fingerprint
ASJC Scopus subject areas
- Computer Networks and Communications
- Information Systems
Cite this
Privacy-preserving source separation for distributed data using independent component analysis. / Imtiaz, Hafiz; Silva, Rogers; Baker, Bradley; Plis, Sergey M.; Sarwate, Anand D.; Calhoun, Vince Daniel.
2016 50th Annual Conference on Information Systems and Sciences, CISS 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 123-127 7460488.Research output: Chapter in Book/Report/Conference proceeding › Conference contribution
}
TY - GEN
T1 - Privacy-preserving source separation for distributed data using independent component analysis
AU - Imtiaz, Hafiz
AU - Silva, Rogers
AU - Baker, Bradley
AU - Plis, Sergey M.
AU - Sarwate, Anand D.
AU - Calhoun, Vince Daniel
PY - 2016/4/26
Y1 - 2016/4/26
N2 - Building good feature representations and learning hidden source models typically requires large sample sizes. In many applications, however, the size of the sample at an individual data holder may not be sufficient. One such application is neuroimaging analyses for mental health disorders - there are many individual research groups, each with a moderate number of subjects. Pooling such data can enable efficient feature learning, but privacy concerns prevent sharing the underlying data. We propose a model for private feature learning in which the data holders share differentially private views of their respective datasets to enable collaborative learning of a joint feature map. We give an example of such an algorithm for independent component analysis (ICA) - a popular blind source separation algorithm used in neuroimaging analyses. Our algorithm is a differentially private version of the recently proposed distributed joint ICA algorithm. We evaluate the performance of this method on simulated functional magnetic resonance imaging (fMRI) data.
AB - Building good feature representations and learning hidden source models typically requires large sample sizes. In many applications, however, the size of the sample at an individual data holder may not be sufficient. One such application is neuroimaging analyses for mental health disorders - there are many individual research groups, each with a moderate number of subjects. Pooling such data can enable efficient feature learning, but privacy concerns prevent sharing the underlying data. We propose a model for private feature learning in which the data holders share differentially private views of their respective datasets to enable collaborative learning of a joint feature map. We give an example of such an algorithm for independent component analysis (ICA) - a popular blind source separation algorithm used in neuroimaging analyses. Our algorithm is a differentially private version of the recently proposed distributed joint ICA algorithm. We evaluate the performance of this method on simulated functional magnetic resonance imaging (fMRI) data.
UR - http://www.scopus.com/inward/record.url?scp=84992406261&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84992406261&partnerID=8YFLogxK
U2 - 10.1109/CISS.2016.7460488
DO - 10.1109/CISS.2016.7460488
M3 - Conference contribution
AN - SCOPUS:84992406261
SP - 123
EP - 127
BT - 2016 50th Annual Conference on Information Systems and Sciences, CISS 2016
PB - Institute of Electrical and Electronics Engineers Inc.
ER -