Linear Layers and Partial Weight Reinitialization for Accelerating Neural Network Convergence

Abstract

We present two new approaches for accelerating the training of a neural network: 1) self-pruning using collapsible linear layers, and 2) mid-training weight reinitialization. By following each nonlinear layer with linear layers, then folding these linear layers into subsequent nonlinear layers after training, we are able to reproduce the benefits of overparameterizing the network, then pruning individual elements after training. We also periodically reinitialize the weights of nonlinear elements that do not improve the networks performance during training, freezing retained weights for several epochs to force the reinitialized weights to accommodate information already learned. Both methods demonstrate substantial gains: the resulting models are simpler than those attained by standard pruning and initialization methods, require fewer computations to train, and are more accurate than networks trained with those methods.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 2018
Accession Number
AD1059350

Entities

People

  • John S. Hyatt
  • Michael S. Lee

Organizations

  • United States Army Research Laboratory

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Artificial Intelligence Software
  • Artificial Neural Networks
  • Boundaries
  • Computer Programming
  • Computer Programs
  • Computers
  • Computing-Computer_Activity
  • Convergence
  • Convolutional Neural Networks
  • Data Compression
  • Data Science
  • Data Set
  • Department Of Defense
  • Digital Data
  • Digital Information
  • Floating Point Operations
  • Image Recognition
  • Images
  • Information Operations
  • Information Processing
  • Information Science
  • Information Systems
  • Machine Learning
  • Mathematical Analysis
  • Mathematics
  • Military Research
  • Neural Networks
  • Signal Processing
  • Standards
  • Training
  • Unmanned Vehicles

Fields of Study

  • Computer science

Readers

  • Adaptive Control and Estimation with Uncertainty in Dynamic Systems.
  • Materials Science
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks