Learning Maneuvers Using Neural Network Models
Abstract
The researchers explored issued involved in implementing robot learning for a challenging dynamic task, using a case study from robot juggling. They used a memory based local modeling approach (locally weighted regression) to represent a learned model of the task to be performed. Statistical tests are given to examine the uncertainty of a model, to optimize its prediction quality, and to deal-with noisy and corrupted data. They developed an exploration algorithm that explicitly deals with prediction accuracy requirements during exploration. Using all these ingredients in combination with methods from optimal control, the robot achieves fast real-time learning of the task within 40 to 100 trials.
Document Details
- Document Type
- Technical Report
- Publication Date
- Aug 07, 1994
- Accession Number
- ADA286470
Entities
People
- Christopher Atkeson
Organizations
- Massachusetts Institute of Technology