Accelerating Imitation Learning in Relational Domains via Transfer by Initialization

Abstract

The problem of learning to mimic a human expert/teacher from training trajectories is called imitation learning. To make the process of teaching easier in this setting, we propose to employ transfer learning (where one learns on a source problem and transfers the knowledge to potentially more complex target problems). We consider multirelational environments such as real-time strategy games and use functional-gradient boosting to capture and transfer the models learned in these environments.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 28, 2013
Accession Number
AD1014298

Entities

People

  • Kristian Kersting
  • Phillip Odom
  • Prasad Tadepalli
  • Saket Joshi
  • Sriraam Natarajan
  • Tushar Khot

Organizations

  • Indiana University

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Air Force
  • Air Force Research Laboratories
  • Algorithms
  • Artificial Intelligence
  • Environment
  • Information Operations
  • Learning
  • Machine Learning
  • Military Research
  • Probability
  • Reinforcement Learning
  • Relational Database Management Systems
  • Supervised Machine Learning
  • Test Beds
  • Training
  • Trajectories

Fields of Study

  • Computer science

Readers

  • Artificial Intelligence
  • Neural Network Machine Learning.