摘要 :
Intensification and diversification are two highly important components of tabu search. At many cases, these two components are also conflicting between them. It is worthy of being studied that how to harmonize this conflict. Aimi...
展开
Intensification and diversification are two highly important components of tabu search. At many cases, these two components are also conflicting between them. It is worthy of being studied that how to harmonize this conflict. Aiming at this problem, a novel adaptive search strategy of intensification and diversification was proposed in this paper. Taking traveling salesman problem as samples, many experiments was investigated. The results shows: this strategy is feasible and effective.
收起
摘要 :
Program understanding is the process of making sense of a complex source code. This process has been considered as computationally difficult and conceptually complex. So far no formal complexity results have been presented, and co...
展开
Program understanding is the process of making sense of a complex source code. This process has been considered as computationally difficult and conceptually complex. So far no formal complexity results have been presented, and conceptual models differ from one researcher to the next. We formally prove that program understanding is NP hard. Furthermore, we show that even a much simpler subproblem remains NP hard. However we do not despair by this result, but rather offer an attractive problem solving model for the program understanding problem. Our model is built on a framework for solving constraint satisfaction problems, or CSPs, which are known to have interesting heuristic solutions. Specifically, we can represent and heuristically address previous and new heuristic approaches to the program understanding problem with both existing and specially designed constraint propagation and search algorithms.
收起
摘要 :
Introduces a generic framework for evolutionary search algorithms (ESAs) and shows how we can use it to imagine the space of all possible ESAs. The idea of the performance landscape for a given search problem is then introduced in...
展开
Introduces a generic framework for evolutionary search algorithms (ESAs) and shows how we can use it to imagine the space of all possible ESAs. The idea of the performance landscape for a given search problem is then introduced in relation to this space of an possible ESAs. The author has previously (1999) argued that search problems can be usefully classified by the most effective ESA on that problem, which translates into the highest peak of the performance landscape. The reason for introducing performance landscapes is to get a better feel for the implications of the "no-free lunch" theorem, together with the notion of real-world problems. These are discussed in relation to performance landscapes. Finally, using the generic framework, a simple generic ESA is described and then used to glean a view of the performance landscape of some NK fitness landscapes (S. Kauffman, 1993), with N=20 and K=2, 5 or 10, and a small real-world problem. With these results, a simple classification of each search problem is made.
收起
摘要 :
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with t...
展开
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with the help of ordering of the operators and searching technique. The problem space could also be viewed as a maze of states and in order to find the goal state we have to do a powerful search on this maze. We show how subgoaling strategy improves search efficiency to get desired goal fast and how it can be incorporated into conventional search methods. In order to improve search we have introduced two-staged scheme of subgoal decomposition and subgoal ordering. This scheme makes use of the guiding power of subgoaling strategies as form of augmented control knowledge in search. The implementation of A* heuristic search algorithm with subgoaling strategy called subgoaling decomposition A* (SDA*) is proposed and examined. The quantitative tradeoff between conventional A* and SDA* in solving some benchmark tasks is investigated to draw the effectiveness of the subgoaling strategies.
收起
摘要 :
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with t...
展开
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with the help of ordering of the operators and searching technique. The problem space could also be viewed as a maze of states and in order to find the goal state we have to do a powerful search on this maze. We show how subgoaling strategy improves search efficiency to get desired goal fast and how it can be incorporated into conventional search methods. In order to improve search we have introduced two-staged scheme of subgoal decomposition and subgoal ordering. This scheme makes use of the guiding power of subgoaling strategies as form of augmented control knowledge in search. The implementation of A* heuristic search algorithm with subgoaling strategy called subgoaling decomposition A* (SDA*) is proposed and examined. The quantitative tradeoff between conventional A* and SDA* in solving some benchmark tasks is investigated to draw the effectiveness of the subgoaling strategies.
收起
摘要 :
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with t...
展开
Problem consists of various states and solving a problem can be considered as searching the space within which these states exits. States are views of the problem and we start at the initial state and aim for the goal state with the help of ordering of the operators and searching technique. The problem space could also be viewed as a maze of states and in order to find the goal state we have to do a powerful search on this maze. We show how subgoaling strategy improves search efficiency to get desired goal fast and how it can be incorporated into conventional search methods. In order to improve search we have introduced two-staged scheme of subgoal decomposition and subgoal ordering. This scheme makes use of the guiding power of subgoaling strategies as form of augmented control knowledge in search. The implementation of A* heuristic search algorithm with subgoaling strategy called subgoaling decomposition A* (SDA*) is proposed and examined. The quantitative tradeoff between conventional A* and SDA* in solving some benchmark tasks is investigated to draw the effectiveness of the subgoaling strategies.
收起
摘要 :
We revisit the problem of searching for a target at an unknown location on a line when given upper and lower bounds on the distance D that separates the initial position of the searcher from the target. Prior to this work, only as...
展开
We revisit the problem of searching for a target at an unknown location on a line when given upper and lower bounds on the distance D that separates the initial position of the searcher from the target. Prior to this work, only asymptotic bounds were known for the optimal competitive ratio achievable by any search strategy in the worst case. We present the first tight bounds on the exact optimal competitive ratio achievable, parametrized in terms of the given range for D, along with an optimal search strategy that achieves this competitive ratio. We prove that this optimal strategy is unique and that it cannot be computed exactly in general. We characterize the conditions under which an optimal strategy can be computed exactly and, when it cannot, we explain how numerical methods can be used efficiently. In addition, we answer several related open questions and we discuss how to generalize these results to m rays, for any m ≥ 2.
收起
摘要 :
Local search (LS) is a widely used, general approach for solving hard combinatorial search problems, such as the graph coloring problem (GCP). One main advantage of this method is that effective heuristics for a problem may lead t...
展开
Local search (LS) is a widely used, general approach for solving hard combinatorial search problems, such as the graph coloring problem (GCP). One main advantage of this method is that effective heuristics for a problem may lead to improvements in solving other problems. Recently, it has been shown that an initial LS algorithm for the Boolean satisfiability problem (SAT) called WalkSAT is extremely effective for random SAT instances. Thus, it is interesting to apply the heuristics in WalkSAT to GCP. This paper proposes a random walk based heuristic, which is inspired by WalkSAT but differs in the tiebreaking mechanism. This new heuristic leads to a new LS algorithm for GCP namely FWLS. The experiments on the DIMACS benchmark show that FWLS finds optimal (or best known) solutions for most instances. Also, when compared to other GCP algorithms, including a greedy one, an LS one and a hybrid one, FWLS exhibits very competitive or better performance.
收起
摘要 :
The maximum diversity problem (MDP) is a classic NP-hard optimization problem with a number of applications. We propose in this work an effective hybrid evolutionary algorithm for MDP called the diversification-driven memetic algo...
展开
The maximum diversity problem (MDP) is a classic NP-hard optimization problem with a number of applications. We propose in this work an effective hybrid evolutionary algorithm for MDP called the diversification-driven memetic algorithm by introducing a diversification mechanism into an existing memetic algorithm. Computational results on 20 representative benchmark instances show that the proposed algorithm outperforms the state-of-the-art MDP algorithms in the literature, indicating the interest of the diversification mechanism within the proposed memetic algorithm.
收起