Scott D. Goodwin

Keep-Best Reproduction: A Local Family Competition Selection Strategy and the Environment it Flourishes in

This paper presents a comparison of two genetic algorithms (GAs) for constrained ordering problems. The first GA uses the standard selection strategy of roulette wheel selection and generational replacement (STDS), while the second GA uses an intermediate selection strategy in addition to STDS. This intermediate selection strategy keeps only the superior offspring and replaces the inferior offspring with the superior parent. We call this selection strategy Keep–Best Reproduction (KBR). The effect of recombination alone, mutation alone and both together are studied. We compare the performance of the different selection strategies and discuss the environment that each selection strategy needs to flourish in. Overall, KBR is found to be the selection strategy of choice. We also present empirical evidence that suggests that KBR is more robust than STDS with regard to operator probabilities and works well with smaller population sizes.

ASERC - A Genetic Sequencing Operator for Asymmetric Permutation Problems

Genetic Algorithms (GAs) have traditionally been designed to work on bitstrings. More recently interest has shifted to the application of GAs to constraint optimization and combinatorial optimization problems. Important for an effective and efficient search is the use of a suitable crossover operator. This paper analyses the performance of six existing crossover operators in the traveling salesman domain. While the edge recombination operator was reported to be the most suitable operator in the TSP domain, our results suggest that this is only true for symmetric TSPs. The problem with edge recombination is that it inverts edges found in the parents. This has no negative effect for the symmetric TSP but can have a substantial effect if the TSP is asymmetric. We propose an edge based crossover operator for the asymmetric TSP and demonstrate its superiority over the traditional edge recombination. Another interesting finding is that order crossover (OX) which has an average performance for symmetric problems, performs very well on asymmetric problems.

Convergence characteristics of keep-best reproduction

This paper presents theoretical convergence characteristics of Keep-Best Reproduction (KBR), a selection strategy for genetic algorithms (GAS). KBR was previously introduced and encouraging results were reported in the traveling salesman domain [16, 181 where KBR was compared with the standard replacement strategy of replacing the two parents by their two children. Here we demonstrate that in a non-operator environment as well as in the ONEMAX domain KBR has the same convergence characteristics as P-tournament selection and elitist recombination (ELR) [13]. We also show how a modification of ELR suggested in [15] can be utilized to tune the selection pressure of KBR. These analytical models are fairly simplistic and cannot accurately model the convergence characteristics in more complex domains where building blocks are correlated, such as the TSP domain. We will give some empirical results of a comparison of KBR and ELR in this domain.

Constraint-Directed Backtracking

We propose a new backtracking algorithm called constraint- directed backtracking (CDBT) for solving general constraint-satisfaction problems (CSPs). CDBT searches for an assignment to variables in a variable set from a given constraint posed on that variable set and appends it to an existing partial solution, in contrast with the naive backtracking (BT) which searches for an assignment of one variable from its domain. In this way, CDBT has a more limited search space and it actually visits fewer nodes than BT. Like BT, CDBT can be improved by incorporating other tree seach techniques such as backjumping or forward checking and consistency techniques such as the ω-consistency algorithm.[/ p]

The Effect of Genetic Operator Probabilities and Selection Strategies on the Performance of a Genetic Algorithm

This paper presents a comparison of two genetic algorithms (GAs) that use different selection strategies. The first GA uses the standard selection strategy of roulette wheel selection and generational replacement (STDS), while the second GA uses an intermediate selection strategy in addition to STDS. Our previous research has shown that this intermediate selection strategy, which we call “Keep-Best Reproduction (KBR)”, found solutions of lower cost for a variety of travelling salesman problems. In this paper, we study the effects of crossover and mutation probabilities on STDS as well as on KBR. We study the effect of recombination alone, mutation alone and both together. We compare the performance of the different selection strategies and discuss the environment that each selection strategy needs to flourish in. Overall, KBR is found to be the selection strategy of choice. We also present empirical evidence that suggests that KBR is more robust than STDS with regard to operator probabilities.

On Dual Encodings for Non-binary Constraint Satisfaction Problems

In [Walsh and Stergiou, 1999] enforcing arc consistency (AC) in the dual encoding was shown to strictly dominate enforcing AC on the hidden or GAC on the original problem. We introduce a dual encoding that requires only a small subset of the original constraints to be stored in extension, while the remaining constraints can be stored intensionally. In this paper we present a theoretical comparison between the pruning achieved by enforcing AC on this dual encoding, versus enforcing GAC and dual arc consistency on the standard encoding. We show how the covering based encoding retains the dominance over enforcing GAC on the original problem, while using less space than the existing dual encoding.

Consistency in general CSPs

In this paper, we introduce a new form of consistency in general constraint satisfaction problems (CSPs), called ω-pairwise-consistency which can be applied to both binary and non-binary constraints. Enforcing ω-pairwise-consistency in a CSP simplifies the problem representation by removing those tuples from the given constraints that will not participate in any solution. Typical CSP solving algorithms can then solve the simplified CSP more efficiently. We show that ω-pairwise-consistency is stronger than some other forms of consistency reported in the literature. We also present an algorithm for achieving ω-pairwise-consistency.

Characterizing Tractable CSPs

In this paper, we introduce the notion of ω-graph as a representative graph for the hypergraph associated with general constraint satisfaction problems (CSPs) and define a new form of consistency called ω-consistency. We identify relationships between the structural property of the ω-graph and the level of ω-consistency that are sufficient to ensure tractability of general CSPs and we prove that the class of tractable CSPs identified here contains the class of tractable CSPs identified with some related conditions reported previously.

Probabilistic Regions of Persistence

Perhaps the most difficult, and certainly the most intensely studied problem in temporal reasoning is the persistence of information-that is, what reasonable inferences can we draw about non-change given partial knowledge of the world and of the changes taking place. Almost all previous work hinges on McCarthys common sense law of inertia (CSLI): things tend not to change. The obvious consequence of adopting this view is that it becomes reasonable to infer that the duration of non-change is arbitrarily long. For instance, a typical inference in systems that appeal to CSLI is that if a person is alive now, the person will remain alive (arbitrarily long) until something happens that results in the persons death.

IT'S ABOUT TIME: AN INTRODUCTION TO THE SPECIAL ISSUE ON TEMPORAL REPRESENTATION AND REASONING

This special issue on temporal representation and reasoning contains seven articles written by the authors of the best papers of the TIME-94 International Workshop on Temporal Representation and Reasoning, which was held in Pensacola, Florida, on May 4, 1994. The workshop’s goal was to bring together researchers concerned with temporal representation and reasoning in a wide variety of areas and promote interaction and cross-fertilization. Consequently, the articles here, while focused on time, address a wide range of issues. The article by Chittaro and Montanari discusses the efficient implementation of Kowalski and Sergot’s event calculus (extended with context dependency). They define a cache-based implementation that moves computational complexity from query processing to update processing and features an absolute improvement in performance because query processing in the event calculus costs much more than update processing in the proposed cached version. Chleq is concerned with the efficient implementation of abduction for temporal reasoning. He defines and implements an abductive procedure based on a constrained resolution principle that efficiently handles ordering relations and can limit hypotheses to ordering relations. Leasure’s article proposes a formal theory of concurrent actions based on the modal logic Z. The theory addresses the qualification, ramification, and frame problems and allows forward and backward reasoning and temporal explanation. It extends Lifschitz and Rabinov’s miracle concept to encompass concurrent actions. Temporal reasoning involving uncertainty about event ordering is the topic of the article by Lin and Dean. They investigate locality in event ordering and causal dependencies. Locality is exploited by using subgoals and abstract events. Their analysis shows a significant performance gain using localized reasoning. Moms et al. formulate a domain-independent framework for reasoning about recurring events and their relations. They propose an ontology of recurrence based on the modeltheoretic structure underlying collective predication using plural noun phrases. Their reasoning framework, which is a specialization of the constraint satisfaction problem framework, allows the manipulation of knowledge stored in a temporal relation network to determine the network’s consistency or to generate scenarios. Provetti is concerned with hypothetical reasoning about actions. He describes how situation calculus cannot represent actual actions while event calculus cannot represent hypothetical actions. He then presents an extention to event calculus capable of representing and reasoning with both actual and hypothetical actions. The resulting formalism is easily implemented as a Prolog program. The article by Tambe and Rosenbloom examines event tracking in a dynamic multiagent environment, namely, air-combat simulation. In addition to monitoring events initiated by other agents, a given agent must be able to make inferences about unobserved events. All this must be handled in an environment of reactive behaviors, continuous interactions, and real-time events. Temporal representation and reasoning has always been and continues to be an important area of research in AI. The approaches have become quite diverse and specialized. Because of