Adaptive functional module selection using machine learning: Framework for intelligent robotics

Martin Lukac, Michitaka Kameyama

Research output: Chapter in Book/Report/Conference proceedingConference contribution

4 Citations (Scopus)

Abstract

In robotics, it is a common problem that for a given task many algorithms are available. For a particular environmental context and some computational constraints some algorithms will perform better and others will perform worse. Consequently, a robot, evolving in a real world environment where both the context and the constraints change in real time, should be able to select in real time algorithms that will provide it with the most accurate world description as well as will allow it to extract the currently most vital information and artifacts. In this paper we propose a machine learning based approach for the real-time selection of computational resources (algorithms) based on both the high level objectives of the robot as well as on the low level environmental requirements (image quality, etc.). The learning mechanism described is using a Genetic Algorithm and the learning method is based on supervised learning; an initial set of algorithms with input data is provided as examples that are used for learning.

Original languageEnglish
Title of host publicationProceedings of the SICE Annual Conference
Pages2480-2483
Number of pages4
Publication statusPublished - 2011
Externally publishedYes
Event50th Annual Conference on Society of Instrument and Control Engineers, SICE 2011 - Tokyo, Japan
Duration: Sep 13 2011Sep 18 2011

Other

Other50th Annual Conference on Society of Instrument and Control Engineers, SICE 2011
CountryJapan
CityTokyo
Period9/13/119/18/11

Fingerprint

Learning systems
Robotics
Robots
Supervised learning
Image quality
Genetic algorithms

Keywords

  • Adaptive Algorithm Selection
  • Intelligent Robotics
  • Machine Learning

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Control and Systems Engineering
  • Computer Science Applications

Cite this

Lukac, M., & Kameyama, M. (2011). Adaptive functional module selection using machine learning: Framework for intelligent robotics. In Proceedings of the SICE Annual Conference (pp. 2480-2483). [6060395]

Adaptive functional module selection using machine learning : Framework for intelligent robotics. / Lukac, Martin; Kameyama, Michitaka.

Proceedings of the SICE Annual Conference. 2011. p. 2480-2483 6060395.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Lukac, M & Kameyama, M 2011, Adaptive functional module selection using machine learning: Framework for intelligent robotics. in Proceedings of the SICE Annual Conference., 6060395, pp. 2480-2483, 50th Annual Conference on Society of Instrument and Control Engineers, SICE 2011, Tokyo, Japan, 9/13/11.
Lukac M, Kameyama M. Adaptive functional module selection using machine learning: Framework for intelligent robotics. In Proceedings of the SICE Annual Conference. 2011. p. 2480-2483. 6060395
Lukac, Martin ; Kameyama, Michitaka. / Adaptive functional module selection using machine learning : Framework for intelligent robotics. Proceedings of the SICE Annual Conference. 2011. pp. 2480-2483
@inproceedings{eb2d1c33bb9b4b22abfbdf875238fbd4,
title = "Adaptive functional module selection using machine learning: Framework for intelligent robotics",
abstract = "In robotics, it is a common problem that for a given task many algorithms are available. For a particular environmental context and some computational constraints some algorithms will perform better and others will perform worse. Consequently, a robot, evolving in a real world environment where both the context and the constraints change in real time, should be able to select in real time algorithms that will provide it with the most accurate world description as well as will allow it to extract the currently most vital information and artifacts. In this paper we propose a machine learning based approach for the real-time selection of computational resources (algorithms) based on both the high level objectives of the robot as well as on the low level environmental requirements (image quality, etc.). The learning mechanism described is using a Genetic Algorithm and the learning method is based on supervised learning; an initial set of algorithms with input data is provided as examples that are used for learning.",
keywords = "Adaptive Algorithm Selection, Intelligent Robotics, Machine Learning",
author = "Martin Lukac and Michitaka Kameyama",
year = "2011",
language = "English",
isbn = "9784907764395",
pages = "2480--2483",
booktitle = "Proceedings of the SICE Annual Conference",

}

TY - GEN

T1 - Adaptive functional module selection using machine learning

T2 - Framework for intelligent robotics

AU - Lukac, Martin

AU - Kameyama, Michitaka

PY - 2011

Y1 - 2011

N2 - In robotics, it is a common problem that for a given task many algorithms are available. For a particular environmental context and some computational constraints some algorithms will perform better and others will perform worse. Consequently, a robot, evolving in a real world environment where both the context and the constraints change in real time, should be able to select in real time algorithms that will provide it with the most accurate world description as well as will allow it to extract the currently most vital information and artifacts. In this paper we propose a machine learning based approach for the real-time selection of computational resources (algorithms) based on both the high level objectives of the robot as well as on the low level environmental requirements (image quality, etc.). The learning mechanism described is using a Genetic Algorithm and the learning method is based on supervised learning; an initial set of algorithms with input data is provided as examples that are used for learning.

AB - In robotics, it is a common problem that for a given task many algorithms are available. For a particular environmental context and some computational constraints some algorithms will perform better and others will perform worse. Consequently, a robot, evolving in a real world environment where both the context and the constraints change in real time, should be able to select in real time algorithms that will provide it with the most accurate world description as well as will allow it to extract the currently most vital information and artifacts. In this paper we propose a machine learning based approach for the real-time selection of computational resources (algorithms) based on both the high level objectives of the robot as well as on the low level environmental requirements (image quality, etc.). The learning mechanism described is using a Genetic Algorithm and the learning method is based on supervised learning; an initial set of algorithms with input data is provided as examples that are used for learning.

KW - Adaptive Algorithm Selection

KW - Intelligent Robotics

KW - Machine Learning

UR - http://www.scopus.com/inward/record.url?scp=81255129380&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=81255129380&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:81255129380

SN - 9784907764395

SP - 2480

EP - 2483

BT - Proceedings of the SICE Annual Conference

ER -