Compressed Sensing of Big Data Networks
Event details
Date | 20.12.2017 |
Hour | 10:00 › 12:00 |
Speaker |
Prof. Alexander Jung, Aalto University Bio: Alexander Jung received the Diplom-Ingenieur (equivalent to MSc) and Dr. techn. (equivalent to Phd) degrees in electrical engineering/signal processing from Technical University Wien, Vienna, Austria in 2008 and 2012, respectively. Between 2012 and 2015 he held Post-Doc positions at ETH Zurich and TU Wien. In 2015 he has been appointed Assitant Professor for Machine Learning at Aalto University, Espoo, Finland. He has received numerous national and international awards, including a best student paper award at the conference IEEE ICASSP 2011 and the "Promotio sub auspiciis Praesidentis rei publicae" (highest academic distinction achievable in Austria). His current research revolves around fundamental limits of and efficient algorithms for machine learning based on massive network-structured datasets (big data over networks). He is co-editor of the special research topic "Compressed Sensing over Complex Networks for Learning from Big Data over Networks" within Frontiers in Applied Mathematics and Statistics. |
Location | |
Category | Conferences - Seminars |
In this talk, we discuss our recent work on developing a theory of compressed sensing for graph signals defined over complex networks. These graph signals represent the label information contained in massive network-structured datasets (big data over networks). By drawing on compressed sensing for ordinary sparse signals, we have introduced the network nullspace property (NNSP) and the network compatibility condition (NCC), which guarantee that certain convex learning methods, including the recently proposed network Lasso, accurately learn clustered graph signals from a small number of signal samples. These conditions involvethe connectivity structure of the underlying network via requiring the existence of certain network flows.
Links
Practical information
- Informed public
- Free
- This event is internal
Organizer
- LTS2 Prof. Pierre Vandergheynst
Contact
- LTS2 Lionel Martin