Works title
A case where a spindly two-layer linear network decisively outperforms any neural network with a fully connected input layer
Authors
[ 1 ] Instytut Informatyki, Wydział Informatyki i Telekomunikacji, Politechnika Poznańska | [ P ] employee
Work ID
r2233_2021
Keywords
PL
- neural networks
- scale invariance
- spindly network
- gradient descent
- lower bounds
Date
21.10.2021
Language
english
Number of pages or volume of work
23
Type of work
research report
License type
CC BY (attribution alone)
Open Access
Yes
System created by Poznań University of Technology
and Poznan Supercomputing and Networking Center
Log in through eKonto to add to SIS