Skip to main navigation
Skip to search
Skip to main content
The Hebrew University of Jerusalem Home
Approve / Request updates on publications
Home
Profiles
Research units
Research output
Prizes
Search by expertise, name or affiliation
Critical percolation as a framework to analyze the training of deep networks
Zohar Ringel
, Rodrigo De Bem
Racah Institute of Physics
Research output
:
Contribution to conference
›
Paper
›
peer-review
1
Scopus citations
Overview
Fingerprint
Fingerprint
Dive into the research topics of 'Critical percolation as a framework to analyze the training of deep networks'. Together they form a unique fingerprint.
Sort by
Weight
Alphabetically
Keyphrases
Cost Function
100%
Deep Network
100%
Critical Percolation
100%
Numerical Analysis
33%
Global Minimum
33%
Euclidean Space
33%
Learning Algorithm
33%
Experiment Analysis
33%
Local Minima
33%
Deep Learning
33%
Planar Graph
33%
CNN Architecture
33%
Classification Task
33%
Topological Classification
33%
Data Topology
33%
Training Experiment
33%
Structured Input
33%
Reachability
33%
Obstacles to Learning
33%
Computer Science
Learning Algorithm
100%
Local Minimum
100%
Planar Graph
100%
Convolutional Neural Network
100%
Classification Task
100%
Deep Learning Method
100%
Mathematics
Cost Function
100%
Numerical Analysis
33%
Euclidean Space
33%
Local Minimum
33%
Rare Event
33%
Input Data
33%
Planar Graph
33%
Deep Learning Method
33%
Convolutional Neural Network
33%
Biochemistry, Genetics and Molecular Biology
Solution and Solubility
100%