Admissible heuristic

About: Admissible heuristic is a research topic. Over the lifetime, 197 publications have been published within this topic receiving 15329 citations. The topic is also known as: admissible heuristics.

Solving the Longest Simple Path Problem with Heuristic Search.

Abstract: Prior approaches for finding the longest simple path (LSP) in a graph used constraints solvers and genetic algorithms In this work, we solve the LSP problem with heuristic search We first introduce several methods for pruning dominated path prefixes Then, we propose several admissible heuristic functions for this problem Experimental results demonstrate the large impact of the proposed heuristics and pruning rules

Expert maintenance systems in telecommunications

Abstract: An admissible heuristic algorithm for generating a network troubleshooting expert system that minimizes the expected troubleshooting cost and learns better troubleshooting techniques during its operation is presented. The algorithm requires three kinds of input: network topology, component failure rate, and relative costs of tests. It uses a modification of Huffman code that is proven to be an admissible heuristic on the A* search. Learning is shown to be facilitated by recomputing the optimal troubleshooting strategy when any of the inputs has changed. The algorithm is implemented in Prolog. >

Symbolic Algorithms for Token Swapping

Abstract: We study different symbolic algorithms to solve two related reconfiguration problems on graphs: the token swapping problem and the permutation routing via matchings problem. Input to both problems is a connected graph with labeled vertices and a token in each vertex. The goal is to move each token to its destination vertex using swap operations. In the token swapping problem, the goal is to find a solution with a minimum number of swaps. In the permutation routing via matchings problem, the goal is to find a solution with a minimum number of steps, where a step is a set of disjoint swaps which can be performed in parallel. First, we present an A* search algorithm. This algorithm can find optimal solutions if used with an admissible heuristic. We also evaluate the use of non-admissible heuristics. In this case, we prove that the result will deviate from an optimum result by at most an even number of swaps. We also present an algorithm based on Boolean satisfiability. We evaluate our methods on a large set of practical benchmarks.

Intelligent control of dynamical systems. I. Robust identification and control of linear time-varying systems

Abstract: The authors propose an artificial-intelligence-based design procedure which addresses the following robust adaptive stabilization problem: given the inputs and outputs of an unknown time-varying system, construct a family of plants one of which is the true system, and obtain a simultaneous stabilizer for the above family. The procedure consists of replacing the classical plant/compensator control loop by a higher level control system consisting of three collaborating expert systems. The procedure is robust, since it utilizes a simultaneous design scheme with a simultaneous identifier. From the point of view of computation, a heuristic scheme is designed, and an admissible heuristic is constructed even for a tree with time-varying cost functions. >

Heuristic algorithm for the resource constrained scheduling problem during high-level synthesis

Abstract: Scheduling is considered the most important task in a high-level synthesis process. A heuristic algorithm based on the A* search to find optimal schedules quickly is presented. This algorithm reduces the computational effort required to obtain the best schedules on a pre-defined datapath by effectively pruning the non-promising search space. The pruning method is accomplished by an admissible heuristic that estimates the schedule length, or the cost, of a search node represented by a partially scheduled data flow graph. The search node with the least cost is considered the most promising candidate and is expanded next, avoiding an exhaustive search of the problem space. When the costs of the candidate search nodes are identical, the A* search is guided by a depth-first search to speed up the computation. Experimental results on several well known benchmarks with varying resource constraints show the effectiveness of the proposed algorithm. Multicycle, pipelined and chaining execution of operations are supported.