Feature-Based Local Policy Reinforcement Learning
This item is restricted to only allow viewing of the metadata.
Title
Feature-Based Local Policy Reinforcement Learning
Author
Feltenberger , David P .
Advisors
Oates , Tim
Program
Computer Science
UMBC Department
Computer Science and Electrical Engineering
Document Type
thesis
Sponsors
University of Maryland , Baltimore County (UMBC)
Keywords
2d images ; artificial intelligence ; growing neural gas ; local policy ; neural network ; reinforcement learning
Date Issued
2009-01-01
Abstract
The problem of learning to control an agent in an arbitrary environment is difficult . In robotics , the standard approach is to hand-code and manually fine-tune a robot's perception of its environment and the actions it should take given its current state . This is both time-consuming and expensive . A better approach is to learn features and action policies without significant manual intervention . This problem is investigated in the context of learning image features to control a fovea position on an image . Using a self- organizing feature map , features are extracted from images . Controllers are then placed at each node and use reinforcement learning to learn how to move a fovea between areas in an image that closely match features in the feature map . Contributions of this work include determining the impact of network parameters (number of nodes , patch size) and sampling methods (random , random walk , structured walk) on learned features , and an understanding of how to perform local control (as opposed to using a monolithic policy as in most RL approaches) based on learned features .
Identifier
10150
Format
application:pdf
Language
en
Collection
UMBC Theses and Disserations .
Rights Statement
This item may be protected under Title 17 of the U.S. Copyright Law. It is made available by UMBC for non-commercial research and education. For permission to publish or reproduce, please see http://library.umbc.edu/speccoll/rightsreproductions.php or contact Special Collections at speccoll(at)umbc.edu.
Source
Feltenberger_umbc_0434M_10150.pdf
Access Rights
Access limited to the UMBC community. Item may possibly be obtained via Interlibrary Loan through a local library, pending author/copyright holder's permission.
Add tags for Feature-Based Local Policy Reinforcement Learning
you wish to report:
Your comment:
Your Name:
...