Fraunhofer Institute for Industrial Mathematics ITWM

Seminar

Deep Learning Seminar / February 25, 2021, 10:00 – 11:00

From Network Pruning to Network Architecture Search: A Unified Framework based on Group Sparsity

Speaker: Avraam Chatzimichailidis (Fraunhofer ITWM, Department High Performance Computing)

Abstract:

Network Pruning aims at reducing the neural network size with an acceptable egradation in the performance. Depending on the granularity, there are weight pruning, kernel pruning and filter pruning. The different pruning objectives are treated in literature separately, and various methods are proposed for each pruning objective. Another important topic in deep learning is network architecture search, which aims at finding a good architecture for the given problem at hand. In this seminar, we show how these seemingly unrelated problems can be effectively tackled under a unified framework based on group sparsity.

Event details

Program

Our speaker Avraam Chatzimichailidis (Fraunhofer ITWM, Department High Performance Computing) presents on the topic »From Network Pruning to Network Architecture Search: A Unified Framework based on Group Sparsity«.

Contact

The seminar is held with the tool MS Teams. Click here on the meeting link. In order to use all functions, it is recommended to download the MS Teams App beforehand.

If you have any questions regarding the organisation, please contact Dominik Straßel (see contacts).

Location

The event takes place digitally.

Date

February 25, 2021, 10:00 – 11:00

Download this event as iCal

Language

English

About the Seminar Series

The Deep Learning Seminar is organized by employees of the field »Data Analysis and Machine Learning« in the department »High Performance Computing«. It is intended to give interested persons an insight into this large field of research and a deeper understanding. Everyone who wants to learn more about Deep Learning, Machine Learning or AI in general is invited - no matter if students, PhD students, professors or software developers. In addition to the employees of our department, interested external speakers can also give a lecture in our seminar series. We also have the opportunity to invite external speakers. We are always open to suggestions, suggestions or requests.

The event usually takes place every Thursday (except for holiday shifts and summer break). The topic of a talk should either come directly from or is relevant to Deep Learning, Machine Learning, Data Analysis or AI, no matter whether it is about a paper, an own project on or an interesting topic. The complexity can range from an overview talk to a special topic.