Flexible computation is a general framework for decision making under limited computational resources. It enables an agent to allocate limited computational resources to maximize its overall performance or utility. In this paper, we present a strategy for flexible computation, which we call iterative state-space reduction. The main ideas are to reduce a problem space that is difficult to search to one that is relatively easy to explore, to use the optimal solution from the reduced space as an approximate solution to the original problem, and to iteratively apply multiple reductions to progressively find better solutions. The basic operation for state-space reduction is heuristic node pruning, which excludes the nodes that do not seem to lead to high-quality solutions from further examination. Iterative state-space reduction can be combined with a state-space search, such as best-first search or depth-first branch-and-bound (DFBnB). The resulting algorithm can run as an anytime algorithm, which provides a tradeoff between solution quality and computation. Based on an analytical modal, we analyze the probability that one iteration of the iterative process finds a solution. Furthermore, by combining it within DFBnB we apply iterative state-space reduction to three combinatorial problems, the maximum Boolean satisfiability, the symmetric TSP and asymmetric TSP. Our experimental results show that iterative state-space reduction is effective and efficient, making DFBnB find better solutions with less computation. 1
|
7715
|
Computers and Intractability: A Guide to the Theory of NP-Completeness
– Garey, Johnson
- 1979
|
|
805
|
Combinatorial Optimization: Algorithms and Complexity
– Papadimitriou, Steiglitz
- 1982
|
|
778
|
A computing procedure for quantification theory
– Davis, Putnam
- 1960
|
|
481
|
An analysis of time-dependent planning
– Dean, Boddy
- 1988
|
|
468
|
Where the really hard problems are
– Cheeseman, Kanefsky, et al.
- 1991
|
|
383
|
Partial constraint satisfaction
– Freuder, Wallace
- 1992
|
|
293
|
Real-time heuristic search
– Korf
- 1990
|
|
280
|
Depth-first iterative-deepening: An optimal admissible tree search
– Korf
- 1985
|
|
273
|
The Traveling Salesman problem
– Lawler, Lenstra
- 1984
|
|
249
|
and easy distributions of SAT problems
– Hard
- 1992
|
|
175
|
Approximating probabilistic inference in Bayesian belief networks is NPhard
– Dagum, Luby
- 1993
|
|
163
|
Do the right thing
– Russell, Wefald
- 1989
|
|
150
|
Reasoning about beliefs and actions under computational resource constraints
– Horvitz
- 1987
|
|
146
|
Deliberation scheduling for problem solving in timeconstrained enviroments
– Boddy, Dean
- 1994
|
|
138
|
The Theory of Branching Processes
– HARRIS
- 1989
|
|
136
|
Constraint-Directed Search: A Case Study of Job-Shop Scheduling
– Fox
- 1987
|
|
126
|
The traveling salesman problem and minimum spanning trees
– Held, Karp
- 1970
|
|
99
|
Optimal Composition of Real-Time Systems
– Zilberstein, Russell
- 1996
|
|
70
|
Systematic and nonsystematic search strategies
– Langley
- 1992
|
|
61
|
Iterative broadening
– Ginsberg, Harvey
- 1992
|
|
42
|
Asymptotic Experimental Analysis for the Held-Karp Traveling Salesman Bound
– Johnson, McGeoch, et al.
- 1996
|
|
30
|
Inductive Learning of Structural Descriptions: Evaluation Criteria and Comparative Review of Selected Methods
– Dietterich, Michalski
- 1981
|
|
27
|
Searching for an optimal path in a tree with random costs
– Karp, Pearl
- 1983
|
|
26
|
Branch and bound methods
– Balas, Toth
- 1985
|
|
26
|
Performance of linear-space search algorithms
– Zhang, Korf
- 1995
|
|
20
|
Some new branching and bounding criteria for the asymmetric traveling salesman problem
– Carpaneto, Toth
- 1980
|
|
20
|
A branch and bound algorithm for the symmetric traveling salesman problem based on the 1-tree relaxation
– Volgenant, Jonker
- 1982
|
|
19
|
Linear assignment problems
– Martello, Toth
- 1987
|
|
18
|
Lecture notes on approximation algorithms
– Motwani
- 1992
|
|
15
|
Anytime heuristic search: First results
– Hansen, Zilberstein, et al.
- 1997
|
|
13
|
Truncated branch-and-bound: A case study on the asymmetric TSP
– Zhang
- 1993
|
|
12
|
Complete anytime beam search
– Zhang
- 1998
|
|
10
|
transformation: exploiting phase transitions to solve combinatorial optimization problems
– Pemberton, Zhang
- 1996
|
|
9
|
An expected-cost analysis of backtracking and non-backtracking algorithms
– McDiarmid, Provan
- 1991
|
|
9
|
Iterative weakening: Optimal and near-optimal policies for the selection of search bias
– Provost
- 1993
|
|
9
|
An average-case analysis of branch-and-bound with applications: Summary of results
– Zhang, Korf
- 1992
|
|
8
|
Probabilistic analysis of tree search
– McDiarmid
- 1990
|
|
5
|
The ARGOS Image Understanding System
– Rubin
- 1978
|
|
5
|
Depth-first vs. best-first search: New results
– Zhang, Korf
- 1993
|
|
4
|
Using branch-and-bound algorithms to obtain suboptimal solutions. Zeitchrift fur
– Ibaraki, Muro, et al.
- 1983
|
|
4
|
Generating Space Telescope Observation Schedules
– Muscettola, Smith, et al.
- 1989
|
|
4
|
An upper bound on the complexity of iterative-deepening-A
– Patrick, Almulla, et al.
- 1992
|
|
2
|
Incremental Random Search Trees
– Korf, Pemberton, et al.
- 1994
|
|
1
|
Large-Vocabulary Speaker-Dependent Continuous Recognition: The Sphinx System
– Lee
- 1988
|
|
1
|
Flexible and approximate computation through state-space reduction
– Zhang
- 1998
|
|
1
|
Truncated and anytime depth-first branch-and-bound: A case study on the asymmetric Traveling Salesman Problem
– Zhang
- 1999
|
|
