ROAIApr 1, 2012

Learning from Humans as an I-POMDP

arXiv:1204.0274v119 citations
Originality Incremental advance
AI Analysis

This work addresses the challenge of interactive learning from humans for AI agents, presenting a novel theoretical framework that is incremental in extending existing POMDP methods to multi-agent settings.

The paper tackles the problem of an agent learning interactively from a human teacher by formulating it as an Interactive Partially Observable Markov Decision Process (I-POMDP), which provides a principled framework for action selection and belief updates while supporting common teacher signals.

The interactive partially observable Markov decision process (I-POMDP) is a recently developed framework which extends the POMDP to the multi-agent setting by including agent models in the state space. This paper argues for formulating the problem of an agent learning interactively from a human teacher as an I-POMDP, where the agent \emph{programming} to be learned is captured by random variables in the agent's state space, all \emph{signals} from the human teacher are treated as observed random variables, and the human teacher, modeled as a distinct agent, is explicitly represented in the agent's state space. The main benefits of this approach are: i. a principled action selection mechanism, ii. a principled belief update mechanism, iii. support for the most common teacher \emph{signals}, and iv. the anticipated production of complex beneficial interactions. The proposed formulation, its benefits, and several open questions are presented.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes