Imperial College London
Alternate

ProfessorMurrayShanahan

Faculty of Engineering, Department of Computing

Professor in Cognitive Robotics
 
 
 
//

Contact

 

+44 (0)20 7594 8262m.shanahan Website

 
 
//

Location

 

407BHuxley BuildingSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@unpublished{Dilokthanakul:2016,
author = {Dilokthanakul, N and Mediano, PAM and Garnelo, M and Lee, MCH and Salimbeni, H and Arulkumaran, K and Shanahan, M},
title = {Deep unsupervised clustering with Gaussian mixture variational autoencoders},
url = {http://arxiv.org/abs/1611.02648v1},
year = {2016}
}
Download

RIS format (EndNote, RefMan)

TY  - UNPB
AB  - We study a variant of the variational autoencoder model with a Gaussian mixture as a prior distribution, with the goal of performing unsupervised clustering through deep generative models. We observe that the standard variational approach in these models is unsuited for unsupervised clustering, and mitigate this problem by leveraging a principled information-theoretic regularisation term known as consistency violation. Adding this term to the standard variational optimisation objective yields networks with both meaningful internal representations and well-defined clusters. We demonstrate the performance of this scheme on synthetic data, MNIST and SVHN, showing that the obtained clusters are distinct, interpretable and result in achieving higher performance on unsupervised clustering classification than previous approaches.
AU  - Dilokthanakul,N
AU  - Mediano,PAM
AU  - Garnelo,M
AU  - Lee,MCH
AU  - Salimbeni,H
AU  - Arulkumaran,K
AU  - Shanahan,M
PY  - 2016///
TI  - Deep unsupervised clustering with Gaussian mixture variational autoencoders
UR  - http://arxiv.org/abs/1611.02648v1
UR  - http://hdl.handle.net/10044/1/42866
ER  -
Download