Active Sequential Learning with Tactile Feedback

Hannes Saal, Jo–Anne Ting, Sethu Vijayakumar
Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, PMLR 9:677-684, 2010.

Abstract

We consider the problem of tactile discrimination, with the goal of estimating an underlying state parameter in a sequential setting. If the data is continuous and high-dimensional, collecting enough representative data samples becomes difficult. We present a framework that uses active learning to help with the sequential gathering of data samples, using information-theoretic criteria to find optimal actions at each time step. We consider two approaches to recursively update the state parameter belief: an analytical Gaussian approximation and a Monte Carlo sampling method. We show how both active frameworks improve convergence, demonstrating results on a real robotic hand-arm system that estimates the viscosity of liquids from tactile feedback data.

Cite this Paper


BibTeX
@InProceedings{pmlr-v9-saal10a, title = {Active Sequential Learning with Tactile Feedback}, author = {Saal, Hannes and Ting, Jo–Anne and Vijayakumar, Sethu}, booktitle = {Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics}, pages = {677--684}, year = {2010}, editor = {Teh, Yee Whye and Titterington, Mike}, volume = {9}, series = {Proceedings of Machine Learning Research}, address = {Chia Laguna Resort, Sardinia, Italy}, month = {13--15 May}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v9/saal10a/saal10a.pdf}, url = { http://proceedings.mlr.press/v9/saal10a.html }, abstract = {We consider the problem of tactile discrimination, with the goal of estimating an underlying state parameter in a sequential setting. If the data is continuous and high-dimensional, collecting enough representative data samples becomes difficult. We present a framework that uses active learning to help with the sequential gathering of data samples, using information-theoretic criteria to find optimal actions at each time step. We consider two approaches to recursively update the state parameter belief: an analytical Gaussian approximation and a Monte Carlo sampling method. We show how both active frameworks improve convergence, demonstrating results on a real robotic hand-arm system that estimates the viscosity of liquids from tactile feedback data.} }
Endnote
%0 Conference Paper %T Active Sequential Learning with Tactile Feedback %A Hannes Saal %A Jo–Anne Ting %A Sethu Vijayakumar %B Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2010 %E Yee Whye Teh %E Mike Titterington %F pmlr-v9-saal10a %I PMLR %P 677--684 %U http://proceedings.mlr.press/v9/saal10a.html %V 9 %X We consider the problem of tactile discrimination, with the goal of estimating an underlying state parameter in a sequential setting. If the data is continuous and high-dimensional, collecting enough representative data samples becomes difficult. We present a framework that uses active learning to help with the sequential gathering of data samples, using information-theoretic criteria to find optimal actions at each time step. We consider two approaches to recursively update the state parameter belief: an analytical Gaussian approximation and a Monte Carlo sampling method. We show how both active frameworks improve convergence, demonstrating results on a real robotic hand-arm system that estimates the viscosity of liquids from tactile feedback data.
RIS
TY - CPAPER TI - Active Sequential Learning with Tactile Feedback AU - Hannes Saal AU - Jo–Anne Ting AU - Sethu Vijayakumar BT - Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics DA - 2010/03/31 ED - Yee Whye Teh ED - Mike Titterington ID - pmlr-v9-saal10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 9 SP - 677 EP - 684 L1 - http://proceedings.mlr.press/v9/saal10a/saal10a.pdf UR - http://proceedings.mlr.press/v9/saal10a.html AB - We consider the problem of tactile discrimination, with the goal of estimating an underlying state parameter in a sequential setting. If the data is continuous and high-dimensional, collecting enough representative data samples becomes difficult. We present a framework that uses active learning to help with the sequential gathering of data samples, using information-theoretic criteria to find optimal actions at each time step. We consider two approaches to recursively update the state parameter belief: an analytical Gaussian approximation and a Monte Carlo sampling method. We show how both active frameworks improve convergence, demonstrating results on a real robotic hand-arm system that estimates the viscosity of liquids from tactile feedback data. ER -
APA
Saal, H., Ting, J. & Vijayakumar, S.. (2010). Active Sequential Learning with Tactile Feedback. Proceedings of the Thirteenth International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 9:677-684 Available from http://proceedings.mlr.press/v9/saal10a.html .

Related Material