A Game Theoretic Framework for Adversarial Classification

Abstract

Many real world applications, ranging from spam filtering to intrusion detection, are facing malicious adversaries who actively transform the objects under their control to avoid detection. Unfortunately, traditional machine learning techniques are insufficient to handle such adversarial problems directly. Adversaries change the dynamics in standard settings where machine learning techniques are designed to excel. They adopt their attacks to deceive the machine learning models built using the past data.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
May 08, 2017
Accession Number
AD1058547

Entities

People

  • Bhavani Thuraisingham
  • Bowei Xi
  • Murat Kantarcıoğlu

Organizations

  • University of Texas at Dallas

Tags

Communities of Interest

  • Autonomy
  • Biomedical
  • Cyber
  • Engineered Resilient Systems
  • Human Systems

DTIC Thesaurus Topics

  • Algorithms
  • Bayesian Networks
  • Computational Science
  • Computer Science
  • Computers
  • Cyber Defense Techniques
  • Data Mining
  • Data Sets
  • Engineering
  • Information Science
  • Intrusion Detection
  • Kernel Functions
  • Machine Learning
  • Probability
  • Students
  • Supervised Machine Learning
  • Training

Fields of Study

  • Computer science

Readers

  • Neural Network Machine Learning.
  • Strategic Security Studies

Technology Areas

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