An adaptive probabilistic graphical model for representing skills in PbD settings

Author

Dindo, Haris ; Schillaci, Guido

Author_Institution

Dipt. di Ing. Inf., Univ. of Palermo, Palermo, Italy

fYear

2010

fDate

2-5 March 2010

Firstpage

89

Lastpage

90

Abstract

Understanding and efficiently representing skills is one of the most important problems in a general Programming by Demonstration (PbD) paradigm. We present Growing Hierarchical Dynamic Bayesian Networks (GHDBN), an adaptive variant of the general DBN model able to learn and to represent complex skills. The structure of the model, in terms of number of states and possible transitions between them, is not needed to be known a priori. Learning in the model is performed incrementally and in an unsupervised manner.

Keywords

automatic programming; belief networks; human-robot interaction; robot programming; unsupervised learning; adaptive probabilistic graphical model; growing hierarchical dynamic Bayesian network; imitation learning; programming by demonstration; skill representation; unsupervised learning; Acceleration; Adaptive systems; Bayesian methods; Clustering algorithms; Collaborative work; Encoding; Graphical models; Hidden Markov models; Human robot interaction; Robot programming; Dynamic Bayesian Network; Growing Hierarchical Dynamic Bayesian Network; Imitation Learning; Incremental Learning; Machine Learning;

fLanguage

English

Publisher

ieee

Conference_Titel

Human-Robot Interaction (HRI), 2010 5th ACM/IEEE International Conference on

Conference_Location

Osaka

Print_ISBN

978-1-4244-4892-0

Electronic_ISBN

978-1-4244-4893-7

Type

conf

DOI

10.1109/HRI.2010.5453257

Filename

5453257