Hartung-Gorre Verlag

Inh.: Dr. Renate Gorre

D-78465 Konstanz

Fon: +49 (0)7533 97227

Fax: +49 (0)7533 97228

www.hartung-gorre.de

S

Series in Communication Theory 

Edited by Helmut Bölcskei

Vol. 16:

 

 

 

 

 

 

 

Michael Tobias Tschannen

 

Unsupervised Learning:

Model-based clustering
and learned compression

 

1st Edition 2019. XVI, 208 pages. € 64,00.

ISBN 978-3-86628-637-5

 

 

 

 

 

 

 

 

 

 

 

 

 

Abstract:

 

This thesis addresses two central tasks prevalent in many modern data processing, storage, and transmission pipelines: Clustering and compression. Specifically, in the first and the second part of this thesis, we study the problems of subspace clustering and random process clustering, respectively. While clustering problems are arguably among the most archetypal problems in unsupervised learning, compression methods are traditionally hand designed. In the third and fourth part of this thesis, we leverage machine learning techniques for compression, a trend that only emerged recently. In more detail, we propose a deep generative model-based framework for lossy data compression on one hand, and we study compression of neural network models for inference on resource-constrained hardware on the other hand.

 

About hte author:

 

Michael Tschannen was born in 1988 in Lenzburg, Switzerland. He obtained the B.Sc. degree in 2012 from EPFL and the M.Sc. degree in 2014 from ETH Zürich, both in Electrical Engineering. He graduated with the Dr. sc. degree from ETH Zürich in 2018. In fall 2017 he was an Applied Scientist Intern at Amazon AI in Palo Alto, CA. His research interests are in machine learning, signal processing, and optimization.

 

Keywords: Subspace clustering; Random processes; Spectral clustering; Lossy compression; Model compression; Generative modeling; Deep neural networks; Generative adversarial networks

 

 

Series in Communication Theory

Hartung-Gorre Verlag Konstanz

Buchbestellungen in Ihrer Buchhandlung oder direkt:

Hartung-Gorre Verlag D-78465  Konstanz // Germany

Telefon: +49 (0) 7533 97227 // Telefax: +49 (0) 7533 97228

http://www.hartung-gorre.de

eMail: verlag@hartung-gorre.de