CSL Master Calendar

Back to Listing

SINE Group Seminar - "Deep Networks from First Principles" Prof. Yi Ma, University of California, Berkeley, CA

Event Type
Seminar/Symposium
Sponsor
SINE Seminar
Location
Join Zoom Meeting https://illinois.zoom.us/j/81195422175?pwd=LzBaOGZCcnFHUTdDcWNRM3NaWHl2UT09 Meeting ID: 811 9542 2175 Password: 422200
Virtual
Date
Nov 30, 2020   4:00 - 5:00 pm  
Contact
Brenda Roy
E-Mail
broy@illinois.edu
Phone
217-244-1663
Views
98
Originating Calendar
CSL SINE Group

Abstract: In this talk, we offer an entirely “white box’’ interpretation of deep (convolutional) networks from the perspective of compression. In particular, we show how modern deep architectures, linear (convolution) operators and nonlinear activations, and parameters of each layer can be derived from the principle of rate reduction (and invariance). All layers, operators, and parameters of the network are explicitly constructed via forward propagation, instead of learned via back propagation. All components of such a network have precise optimization, geometric, and statistical meaning. There are also several nice surprises from this principled approach that shed new light on fundamental relationships between forward (optimization) and backward (variation) propagation, between invariance and sparsity, and between deep networks and Fourier transform. 

link for robots only