Censi, Andrea and Murray, Richard M. (2011) A group-theoretic approach to formalizing bootstrapping problems. Caltech , Pasadena, CA. (Unpublished) http://resolver.caltech.edu/CaltechCDSTR:2011.005
See Usage Policy.
Use this Persistent URL to link to this item: http://resolver.caltech.edu/CaltechCDSTR:2011.005
The bootstrapping problem consists in designing agents that learn a model of themselves and the world, and utilize it to achieve useful tasks. It is different from other learning problems as the agent starts with uninterpreted observations and commands, and with minimal prior information about the world. In this paper, we give a mathematical formalization of this aspect of the problem. We argue that the vague constraint of having "no prior information" can be recast as a precise algebraic condition on the agent: that its behavior is invariant to particular classes of nuisances on the world, which we show can be well represented by actions of groups (diffeomorphisms, permutations, linear transformations) on observations and commands. We then introduce the class of bilinear gradient dynamics sensors (BGDS) as a candidate for learning generic robotic sensorimotor cascades. We show how framing the problem as rejection of group nuisances allows a compact and modular analysis of typical preprocessing stages, such as learning the topology of the sensors. We demonstrate learning and using such models on real-world range-finder and camera data from publicly available datasets.
|Item Type:||Report or Paper (Technical Report)|
|Group:||Control and Dynamical Systems Technical Reports|
|Usage Policy:||You are granted permission for individual, educational, research and non-commercial reproduction, distribution, display and performance of this work in any format.|
|Deposited By:||Imported from CaltechCDSTR|
|Deposited On:||18 Apr 2011|
|Last Modified:||18 Mar 2015 22:57|
Repository Staff Only: item control page