Measuring and Comparing Robustness of ML Algorithms Under Adversarial Attack

Abstract

A machine learning algorithm can be evaluated for robustness against any number of different types of attacks. We consider attacks that seek to manipulate the training and/or testing data inputs to a machine learning algorithm. Specifically, we do not consider physical attacks on machines hosting the algorithm.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 01, 2017
Accession Number
AD1088314

Entities

People

  • Eliezer Kanal

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Algorithms
  • Artificial Intelligence
  • Denial Of Service Attack
  • Engineering
  • Guarantees
  • Learning
  • Machine Learning
  • Materials
  • Neural Networks
  • Reliability
  • Software Development
  • Supervised Machine Learning
  • Training
  • Universities
  • Unmanned Vehicles

Fields of Study

  • Computer science

Readers

  • Neural Network Machine Learning.
  • Political Violence and Terrorism Studies.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks