NETRIUS Deep Chaos

Abstract

Artificial intelligence, and deep learning in particular, is increasingly becoming a critical component in our national infrastructure. It is imperative that techniques used to train these models be consistent in their predictions. Fooling these systems with small perturbations to the input data is a well-known problem. Identifying and characterizing regions of sensitivity is less understood in the Al domain, but well studied in nonlinear dynamics mathematics. We seek to investigate whether deep neural networks are chaotic, and, if so, can they be stabilized?

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 20, 2021
Accession Number
AD1153246

Entities

People

  • Mohammed Eslami

Tags

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algebraic Topology
  • Algorithms
  • Artificial Intelligence
  • Artificial Intelligence Computing
  • Artificial Intelligence Software
  • Convolutional Neural Networks
  • Deep Learning
  • Department Of Defense
  • Information Processing
  • Information Science
  • Information Systems
  • Neural Networks
  • Nonlinear Dynamics
  • Scientific Research
  • Topology

Readers

  • Control Systems Engineering.
  • Military History / Militaries and War Studies
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - DoD AI Strategy
  • AI & ML - Neural Networks