東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Teaching old dogs new tricks: Increm...

Grollman, Daniel H.

Linked to FindBook

Google Book

Amazon

博客來

Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration./
Author:	Grollman, Daniel H.
Description:	141 p.
Notes:	Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 6871.
Contained By:	Dissertation Abstracts International71-11B.
Subject:	Engineering, Robotics. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3430184
ISBN:	9781124302898

Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration.
Grollman, Daniel H.

Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration. - 141 p.

Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 6871.

Thesis (Ph.D.)--Brown University, 2010.

We consider autonomous robots as having associated control policies that determine their actions in response to perceptions of the environment. Often, these controllers are explicitly transferred from a human via programmatic description or physical instantiation. Alternatively, Robot Learning from Demonstration (RLfD) can enable a robot to learn a policy from observing only demonstrations of the task itself. We focus on interactive, teleoperative teaching, where the user manually controls the robot and provides demonstrations while receiving learner feedback. With regression, the collected perception-actuation pairs are used to directly estimate the underlying policy mapping.

ISBN: 9781124302898Subjects--Topical Terms:

1018454
Engineering, Robotics.

Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration.
LDR:03314nam 2200301 4500 001 1405161
005 20111206130423.5
008 130515s2010 ||||||||||||||||| ||eng d
020 $a 9781124302898
035 $a (UMI)AAI3430184
035 $a AAI3430184
040 $a UMI $c UMI
100 1 $a Grollman, Daniel H. $3 1684519
245 1 0 $a Teaching old dogs new tricks: Incremental multimap regression for interactive robot learning from demonstration.
300 $a 141 p.
500 $a Source: Dissertation Abstracts International, Volume: 71-11, Section: B, page: 6871.
500 $a Adviser: Odest Chadwicke Jenkins.
502 $a Thesis (Ph.D.)--Brown University, 2010.
520 $a We consider autonomous robots as having associated control policies that determine their actions in response to perceptions of the environment. Often, these controllers are explicitly transferred from a human via programmatic description or physical instantiation. Alternatively, Robot Learning from Demonstration (RLfD) can enable a robot to learn a policy from observing only demonstrations of the task itself. We focus on interactive, teleoperative teaching, where the user manually controls the robot and provides demonstrations while receiving learner feedback. With regression, the collected perception-actuation pairs are used to directly estimate the underlying policy mapping.
520 $a This dissertation contributes an RLfD methodology for interactive, mixed-initiative learning of unknown tasks. The goal of the technique is to enable users to implicitly instantiate autonomous robot controllers that perform desired tasks as well as the demonstrator, as measured by task-specific metrics. With standard regression techniques, we show that such "on-par" learning is restricted to policies typified by a many-to-one mapping (a unimap) from perception to actuation. Thus, controllers representable as multi-state Finite State Machines (FSMs) and that exhibit a one-to-many mapping (a multimap) cannot be learnt. To be able to do so we must address the three issues of model selection (how many subtasks or FSM states), policy learning (for each subtask), and transitioning (between subtasks). Previous work in RLfD has assumed knowledge of the task decomposition and learned the subtask policies or the transitions between them in isolation.
520 $a We instead address both model selection and policy learning simultaneously. Our presented technique uses an infinite mixture of experts and treats the multimap data from an FSM controller as being generated from overlapping unimaps. The algorithm automatically determines the number of unimap experts (model selection) and learns a unimap for each one (policy learning). On data from both synthetic and robot soccer multimaps we show that the discovered subtasks can be used (switched between) to reperform the original task. While not at the same level of skill as the demonstrator, the resulting approximations represent significant improvement over ones for the same tasks learned with unimap regression.
590 $a School code: 0024.
650 4 $a Engineering, Robotics. $3 1018454
650 4 $a Artificial Intelligence. $3 769149
690 $a 0771
690 $a 0800
710 2 $a Brown University. $3 766761
773 0 $t Dissertation Abstracts International $g 71-11B.
790 1 0 $a Jenkins, Odest Chadwicke, $e advisor
790 $a 0024
791 $a Ph.D.
792 $a 2010
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3430184