Heuristic Search Procedure Research Articles

Watershed Land-Use Planning under Uncertainty

A model of watershed land-use planning is formulated that improves on existing models by recognizing that land-use decisions have uncertain outcomes and that land uses change over time. Implications of recognizing the distinction between land-use decisions and their uncertain outcomes are discussed. The land-use changes are modelled using a Markov process. Because of the computational difficulties in determining the return associated with the complete range of possible decision sets, a heuristic technique is required. A heuristic search procedure based on stochastic dynamic programming is described.

A general branch and bound formulation for understanding and synthesizing and/or tree search procedures

Citing the confusing statements in the AI literature concerning the relationship between branch and bound (B&B) and heuristic search procedures were present a simple and general formulation of B&B which should help dispel much of the confusion. We illustrate the utility of the formulation by showing that through it some apparently very different algorithms for searching And/Or trees reveal the specific nature of their similarities and differences. In addition to giving new insights into the relationships among some AI search algorithms, the general formulation also provides suggestions on how existing search procedures may be varied to obtain new algorithms.

Optimization problems in computer network design

This paper examines optimization problems associated with computer network design and routing. The emphasis is focused on modifying conventional heuristic search procedures to enhance their solution optimization capability. The techniques discussed include fuzzy heuristics, adaptive compositions of heuristics and suboptimum algorithms based on dynamic programming concepts. The discussion is illustrated with examples and experimental computer implementation results. The results obtained demonstrate that the application of the above methods improves solutions derived by means of conventional heuristic procedures and brings such solutions closer to optimum ones.

Fuzzy-PLANNER: Reasoning with Inexact Concepts in a Procedural Problem-Solving Language

Abstract All contemporary deductive problem-solving paradigms deal with a world in which assertions are true (false) and action-rules valid (invalid). This simplified situation is inadequate for realistic applications which include inexact information. This article describes a precise computationally specific method for coupling two different many-valued logics with a procedural problem-solving system (micro-PLANNER). Solutions to deductive problems can be found which meet specific criteria of validity. This particular scheme enables the system to dynamically compute the truth-value of a subgoal during the search process. Thus, the validity of a subgoal may be used to direct the heuristic search procedure. Fuzzy-PLANNER is a promising medium for experimenting with different many-valued logics to find the ones most appropriate for different problem domains. However, the notions elaborated here are relevant to any procedural problem-solving language.

Breadth-first search: Some surprising results

Although breadth-first search procedures cannot explore truly large search spaces, actual implementations of such procedures can result in surprisingly powerful problem-solvers that outperform more sophisticated heuristic search procedures. We describe two breadth-first search procedures. The first one, S&R, proves theorems from Principia of Whitehead and Russell, and is compared to two versions of the Logic Theorist. Previous estimates of the size of the search space are significantly reduced. When theorems are proved in an optimal order, this order differs markedly from that found in Principia, while more general theorems than those of Principia are often found. The second system, S&M, adapts breadth-first search to locally infinite search spaces in systems of rewriting rules. S&M is compared extensively to the heuristic theorem-prover of Quinlan and Hunt, and to some other theorem provers.

Fuzzy Planner: Reasoning with Inexact Concepts in a Procedural Problem-Solving Language

Abstract All contemporary deductive problem-solving paradigms deal with a world in which assertions are true (false) and action-rules valid (invalid). This simplified situation is inadequate for realistic applications which include inexact information. This report describes a precise computationally specific method for coupling two different many-valued logics with a procedural problem-solving system (micro-PLANNER). Solutions to deductive problems can be found which meet specific criteria of validity. This particular scheme enables the system to dynamically compute the truth-value of a subgoal during the search process. Thus, the validity of a subgoal may be used to direct the heuristic search procedure. Fuzzy PLANNER is a promising medium for experimenting with different many-valued logics to find the ones most appropriate for different problem domains. However, the notions elaborated here are relevent to any procedural problem-solving language.