Planning under uncertainty is a central problem in the study of automated sequential decision making, and has been addressed by researchers in many different fields, including AI planning, decision analysis, operations research, control theory and economics. While the assumptions and perspectives adopted in these areas often differ in substantial ways, many planning problems of interest to researchers in these fields can be modeled as Markov decision processes (MDPs) and analyzed using the techniques of decision theory. This paper presents an overview and synthesis of MDP-related methods, showing how they provide a unifying framework for modeling many classes of planning problems studied in AI. It also describes structural properties of MDPs that, when exhibited by particular classes of problems, can be exploited in the construction of optimal or approximately optimal policies or plans. Planning problems commonly possess structure in the reward and value functions used to de...

Bucket elimination is an algorithmic framework that generalizes dynamic programming to accommodate many problem-solving and reasoning tasks. Algorithms such as directional-resolution for propositional satisfiability, adaptive-consistency for constraint satisfaction, Fourier and Gaussian elimination for solving linear equalities and inequalities, and dynamic programming for combinatorial optimization, can all be accommodated within the bucket elimination framework. Many probabilistic inference tasks can likewise be expressed as bucket-elimination algorithms. These include: belief updating, finding the most probable explanation, and expected utility maximization. These algorithms share the same performance guarantees; all are time and space exponential in the inducedwidth of the problem's interaction graph. While elimination strategies have extensive demands on memory, a contrasting class of algorithms called "conditioning search" require only linear space. Algorithms in this class split a problem into subproblems by instantiating a subset of variables, called a conditioning set, or a cutset. Typical examples of conditioning search algorithms are: backtracking (in constraint satisfaction), and branch and bound (for combinatorial optimization). The paper presents the bucket-elimination framework as a unifying theme across probabilistic and deterministic reasoning tasks and show how conditioning search can be augmented to systematically trade space for time.

Markov decision processes(MDPs) have proven to be popular models for decision-theoretic planning, but standard dynamic programming algorithms for solving MDPs rely on explicit, state-based specifications and computations. To alleviate the combinatorial problems associated with such methods, we propose new representational and computational techniques for MDPs that exploit certain types of problem structure. We use dynamic Bayesian networks (with decision trees representing the local families of conditional probability distributions) to represent stochastic actions in an MDP, together with a decision-tree representation of rewards. Based on this representation, we develop versions of standard dynamic programming algorithms that directly manipulate decision-tree representations of policies and value functions. This generally obviates the need for state-by-state computation, aggregating states at the leaves of these trees and requiring computations only for each aggregate state. The key to these algorithms is a decision-theoretic generalization of classic regression analysis, in which we determine the features relevant to predicting expected value. We demonstrate the method empirically on several planning problems,

Probabilistic independence can dramatically simplify the task of eliciting, representing, and computing with probabilities in large domains. A key technique in achieving these benefits is the idea of graphical modeling. We survey existing notions of independence for utility functions in a multi-attribute space, and suggest that these can be used to achieve similar advantages. Our new results concern conditional additive independence, which we show always has a perfect representation as separation in an undirected graph (a Markov network). Conditional additive independencies entail a particular functional form for the utility function that is analogous to a product decomposition of a probability function, and confers analogous benefits. This functional form has been utilized in the Bayesian network and influence diagram literature, but generally without an explanation in terms of independence. The functional form yields a decomposition of the utility function that can greatly speed up e...

This paper addresses the problem of planning under uncertainty in large Markov Decision Processes (MDPs). Factored MDPs represent a complex state space using state variables and the transition model using a dynamic Bayesian network. This representation often allows an exponential reduction in the representation size of structured MDPs, but the complexity of exact solution algorithms for such MDPs can grow exponentially in the representation size. In this paper, we present two approximate solution algorithms that exploit structure in factored MDPs. Both use an approximate value function represented as a linear combination of basis functions, where each basis function involves only a small subset of the domain variables. A key contribution of this paper is that it shows how the basic operations of both algorithms can be performed efficiently in closed form, by exploiting both additive and context-specific structure in a factored MDP. A central element of our algorithms is a novel linear program decomposition technique, analogous to variable elimination in Bayesian networks, which reduces an exponentially large LP to a provably equivalent, polynomial-sized one. One algorithm uses approximate linear programming, and the second approximate dynamic programming. Our dynamic programming algorithm is novel in that it uses an approximation based on max-norm, a technique that more directly minimizes the terms that appear in error bounds for approximate MDP algorithms. We provide experimental results on problems with over 10^40 states, demonstrating a promising indication of the scalability of our approach, and compare our algorithm to an existing state-of-the-art approach, showing, in some problems, exponential gains in computation time.

The problem of driving an autonomous vehicle in normal traffic engages many areas of AI research and has substantial economic significance. We describe work in progress on a new approach to this problem that uses a decision-theoretic architecture using dynamic probabilistic networks. The architecture provides a sound solution to the problems of sensor noise, sensor failure, and uncertainty about the behavior of other vehicles and about the effects of one's own actions. We report on advances in the theory of inference and decision making in dynamic, partially observable domains. Our approach has been implemented in a simulation system, and the autonomous vehicle successfully negotiates a variety of difficult situations. 1 The BAT Project Several government agencies and corporations in Europe, Japan, and the US are currently undertaking research in IVHS (Intelligent Vehicle and Highway Systems) with the aim of substantially reducing congestion and accidents, which cost $500 billion/year...

The problem of planning under uncertainty has been addressed by researchers in many different fields, adopting rather different perspectives on the problem. Unfortunately, these researchers are not always aware of the relationships among these different problem formulations, often resulting in confusion and duplicated effort. Many probabilistic planning or decision making problems can be characterized as a class of Markov decision processes that allow for significant compression in representing the underlying system dynamics. It is for this class of problems that we as experts in intensional representations are advantageously positioned to contribute efficient solution methods. This paper provides a general characterization of the representational requirements for this class of problems, and we describe how to achieve computational leverage using representations that make different types of dependency information explicit. Keywords: decision-theoretic planning, action repr...

We describe the development of a monitoring system which uses sensor observation data about discrete events to construct dynamically a probabilistic model of the world. This model is a Bayesian network incorporating temporal aspects, which we call a Dynamic Belief Network; it is used to reason under uncertainty about both the causes and consequences of the events being monitored. The basic dynamic construction of the network is data-driven. However the model construction process combines sensor data about events with externally provided information about agents' behaviour, and knowledge already contained within the model, to control the size and complexity of the network. This means that both the network structure within a time interval, and the amount of history and detail maintained, can vary over time. We illustrate the system with the example domain of monitoring robot vehicles and people in a restricted dynamic environment using light-beam sensor data. In addition to presenting a ...

Partially observable Markov decision processes (POMDPs) provide a natural and principled framework to model a wide range of sequential decision making problems under uncertainty. To date, the use of POMDPs in real-world problems has been limited by the poor scalability of existing solution algorithms, which can only solve problems with up to ten thousand states. In fact, the complexity of finding an optimal policy for a finite-horizon discrete POMDP is PSPACE-complete. In practice, two important sources of intractability plague most solution algorithms: large policy spaces and large state spaces. On the other hand,

Probabilistic independence has proved to be a fundamental tool that can dramatically simplify the task of eliciting, representing, and computing with probabilities. We advance the position that notions of utility independence can serve similar functions when reasoning about preferences and utilities during decision making.