A Modal Logic for the Decision-Theoretic Projection Problem

We present a decidable logic in which queries can be posed about (i) the degree of belief in a propositional sentence after an arbitrary finite number of actions and observations and (ii) the utility of a finite sequence of actions after a number of actions and observations. Another contribution of this work is that a POMDP model specification is allowed to be partial or incomplete with no restriction on the lack of information specified for the model. The model may even contain information about non-initial beliefs. Essentially, entailment of arbitrary queries (expressible in the language) can be answered. A sound, complete and terminating decision procedure is provided.

[1]  Gerhard Lakemeyer,et al.  On the Logical Specification of Probabilistic Transition Models , 2013 .

[2]  Eyal Amir,et al.  First-Order Logical Filtering , 2005, IJCAI.

[3]  Zhengzhu Feng,et al.  Dynamic Programming for POMDPs Using a Factored State Representation , 2000, AIPS.

[4]  Emad Saad,et al.  Probabilistic Reasoning by SAT Solvers , 2009, ECSQARU.

[5]  Craig Boutilier,et al.  Decision-Theoretic, High-Level Agent Programming in the Situation Calculus , 2000, AAAI/IAAI.

[6]  George E. Monahan,et al.  A Survey of Partially Observable Markov Decision Processes: Theory, Models, and Algorithms , 2007 .

[7]  Toniann Pitassi,et al.  Stochastic Boolean Satisfiability , 2001, Journal of Automated Reasoning.

[8]  Blai Bonet.  High-Level Planning and Control with Incomplete Information Using POMDPs Hdctor Geffner and , 2003 .

[9]  David Poole,et al.  Decision Theory, the Situation Calculus and Conditional Plans , 1998, Electron. Trans. Artif. Intell..

[10]  Pierre Lison.  Towards Relational POMDPs for Adaptive Dialogue Management , 2010, ACL.

[11]  Edward J. Sondik,et al.  The Optimal Control of Partially Observable Markov Processes over a Finite Horizon , 1973, Oper. Res..

[12]  Gavin Rens,et al.  Formalisms for agents reasoning with stochastic actions and perceptions. , 2014 .

[13]  Scott Sanner,et al.  Symbolic Dynamic Programming for First-order POMDPs , 2010, AAAI.

[14]  Bengt Jonsson,et al.  A logic for reasoning about time and reliability , 1990, Formal Aspects of Computing.

[15]  Gerhard Lakemeyer,et al.  A Logic for Specifying Stochastic Actions and Observations , 2014, FoIKS.

[16]  J. McCarthy.  Situations, Actions, and Causal Laws , 1963 .

[17]  Roni Khardon,et al.  Relational Partially Observable MDPs , 2010, AAAI.

[18]  John McCarthy,et al.  SOME PHILOSOPHICAL PROBLEMS FROM THE STANDPOINT OF ARTI CIAL INTELLIGENCE , 1987 .

[19]  Roni Khardon,et al.  First Order Decision Diagrams for Relational MDPs , 2007, IJCAI.

[20]  Gerhard Lakemeyer,et al.  SLAP: Specification logic of actions with probability , 2014, J. Appl. Log..

[21]  Gerhard Lakemeyer,et al.  Situations, Si! Situation Terms, No! , 2004, KR.

[22]  Joelle Pineau,et al.  A Bayesian Approach for Learning and Planning in Partially Observable Markov Decision Processes , 2011, J. Mach. Learn. Res..

[23]  Marta Kwiatkowska,et al.  Advances and challenges of probabilistic model checking , 2010, 2010 48th Annual Allerton Conference on Communication, Control, and Computing (Allerton).

[24]  Chenggang Wang,et al.  Planning with POMDPs using a compact, logic-based representation , 2005, 17th IEEE International Conference on Tools with Artificial Intelligence (ICTAI'05).

[25]  Craig Boutilier,et al.  Computing Optimal Policies for Partially Observable Decision Processes Using Compact Representations , 1996, AAAI/IAAI, Vol. 2.

[26]  Thomas Lukasiewicz,et al.  Reasoning about actions with sensing under qualitative and probabilistic uncertainty , 2004, TOCL.

[27]  Gerhard Lakemeyer,et al.  ESP: A Logic of Only-Knowing, Noisy Sensing and Acting , 2007, AAAI.