LPARX is a software development tool for implementing dynamic, irregular scientific applications, such as multilevel multilevel finite difference methods and particle methods, on high performance MIMD parallel architectures. It supports coarse grain data parallelism and gives the application complete control over specifying arbitrary block decompositions. LPARX provides structural abstraction, representing data decompositions as first-class objects that can be manipulated and modified at run-time. LPARX, implemented as a C++ class library, is currently running on diverse MIMD platforms, including the Intel Paragon, Cray C-90, IBM SP2, and networks of workstations running under PVM. Software may be developed and debugged on a single processor workstation. 1 Introduction An outstanding problem in scientific computation is how to manage the complexity of converting mathematical descriptions of dynamic, irregular numerical algorithms into high performance applications software. Non-unifo...

This paper presents an approach to combine competent crossover and mutation operators via probabilistic model building. Both operators are based on the probabilistic model building procedure of the extended compact genetic algorithm (eCGA). The model sampling procedure of eCGA, which mimics the behavior of an idealized recombination— where the building blocks (BBs) are exchanged without disruption—is used as the competent crossover operator. On the other hand, a recently proposed BB-wise mutation operator—which uses the BB partition information to perform local search in the BB space—is used as the competent mutation operator. The resulting algorithm, called hybrid extended compact genetic algorithm (heCGA), makes use of the problem decomposition information for (1) effective recombination of BBs and (2) effective local search in the BB neighborhood. The proposed approach is tested on different problems that combine the core of three well known problem difficulty dimensions: deception, scaling, and noise. The results show that, in the absence of domain knowledge, the hybrid approach is more robust than either single-operatorbased approach.

This paper examines the effects of relaxed synchronization on both the numerical and parallel efficiency of parallel genetic algorithms (GAs). We describe a coarse-grain geographically structured parallel genetic algorithm. Our experiments provide preliminary evidence that asynchronous versions of these algorithms have a lower run time than synchronous GAs. Our analysis shows that this improvement is due to (1) decreased synchronization costs and (2) high numerical efficiency (e.g. fewer function evaluations) for the asynchronous GAs. This analysis includes a critique of the utility of traditional parallel performance measures for parallel GAs. 1. Introduction Genetic algorithms (GAs) are stochastic search algorithms that have been successfully applied to a variety of optimization problems [5]. Unlike most other optimization procedures, GAs maintain a population of individuals (set of solutions) that are competitively selected to generate new candidates for the global optima. Parallel...

Molecular docking software makes computational predictions of the interaction of molecules. This can be useful, for example, in evaluating the binding of candidate drug molecules to a target molecule from a virus. In the Autodock docking software (Morris et al. 1996), a physical model is used to evaluate the energy of candidate docked configurations, and heuristic search is used to minimize this energy. Previous versions of Autodock used simulated annealing to do this heuristic search. We evaluate the use of the genetic algorithm with local search in Autodock. We investigate several GA-local search (GA-LS) hybrids and compare results with those obtained from simulated annealing. This comparison is done in terms of optimization success, and absolute success in finding the true physical docked configuration. We use these results to test the energy function and evaluate the success of the application. 1 THE DOCKING PROBLEM When two molecules are in close proximity, it can be energeticall...

The SEARCH (Search Envisioned As Relation & Class Hierarchizing) framework developed elsewhere (Kargupta, 1995) offered an alternate perspective toward blackbox optimization (BBO)---optimization in presence of little domain knowledge. The SEARCH framework investigated the conditions essential for transcending the limits of random enumerative search using a framework developed in terms of relations, classes and partial ordering. This paper presents a summary of some of the main results of that work. A closed form bound on the sample complexity in terms of the cardinality of the relation space, class space, desired quality of the solution and the reliability is presented. The two primary lessons of this work are, a BBO (1) must search for appropriate relations and (2) can only solve the so called class of order-k delineable problems in polynomial sample complexity. These results are applicable to any blackbox search algorithms, including evolutionary optimization techniques. 1 Introducti...

s: Jan. 1995 { Sep. 1998 ACM: ACM Guide to Computing Literature: 1979 - 1993/4 BA: Biological Abstracts: July 1996 - Aug. 1998 CA: Computer Abstracts: Jan. 1993 { Feb. 1995 CCA: Computer & Control Abstracts: Jan. 1992 { Apr. 1998 (except May-95) ChA: Chemical Abstracts: Jan. 1997 - Dec. 1998 CTI: Current Technology Index Jan./Feb. 1993 { Jan./Feb. 1994 DAI: Dissertation Abstracts International: Vol. 53 No. 1 { Vol. 56 No. 10 (Apr. 1996) EEA: Electrical & Electronics Abstracts: Jan. 1991 { Apr. 1998 EI A: The Engineering Index Annual: 1987 - 1992 EI M: The Engineering Index Monthly: Jan. 1993 { Apr. 1998 (except May 1997) N: Scientic and Technical Aerospace Reports: Jan. 1993 - Dec. 1995 (except Oct. 1995) P: Index to Scientic & Technical Proceedings: Jan. 1986 { May 1998 (except Nov. 1994) PA: Physics Abstracts: Jan. 1997 { Sep. 1998 1.1 Your contributions erroneous or missing? The bibliography database is updated on a regular basis and certain...

Many frustrating experiences have been encountered when the training of neural networks by local search methods becomes stagnant at local optima. This calls for the development of more satisfactory search methods such as evolutionary search. However, training by evolutionary search can require a long computation time. In certain situations, using Lamarckian evolution, local search and evolutionary search can complement each other to yield a better training algorithm. This paper demonstrates the potential of this evolutionary-learning synergy by applying it to train recurrent neural networks in an attempt to resolve a long-term dependency problem and the inverted pendulum problem. This work also aims at investigating the interaction between local search and evolutionary search when they are combined. It is found that the combinations are particularly efficient when the local search is simple. In the case where no teacher signal is available for the local search to learn the desired task...

This paper proposes a hybrid optimization algorithm which combines the efforts of local search (individual learning) and cellular genetic algorithms (GAs) for training recurrent neural networks (RNNs). Each weight of an RNN is encoded as a floating point number, and a concatenation of the numbers forms a chromosome. Reproduction takes place locally in a square grid with each grid point representing a chromosome. Two approaches, Lamarckian and Baldwinian mechanisms, for combining cellular GAs and learning have been compared. Different hill-climbing algorithms are incorporated into the cellular GAs as learning methods. These include the real-time recurrent learning (RTRL) and its simplified versions, and the delta rule. The RTRL algorithm has been successively simplified by freezing some of the weights to form simplified versions. The delta rule, which is the simplest form of learning, has been implemented by considering the RNNs as feedforward networks during learning. The hybrid algori...

. This work studies the interaction of evolution and learning. It starts from the coevolutionary genetic algorithm (CGA) introduced earlier. Two techniques - life-time fitness evaluation (LTFE) and predator-prey coevolution - boost the genetic search of a CGA. The partial but continuous nature of LTFE allows for an elegant incorporation of life-time learning (LTL) within CGAs. This way, not only the genetic search but also the LTL component focuses on "not yet solved" problems. The performance of the new algorithm is compared with various other algorithms. 1 Introduction The combination of evolutionary learning and life-time learning (LTL) in genetic algorithms (GAs) is an active field of research nowadays. The rise of interest in the combination of both types of learning has several reasons. First of all, it is clear that nature combines both types of learning. Hence, it is of interest to gather a better understanding of the interaction between, and the advantages of, both types of l...

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii I Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 A. Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1. The #arti#cial life bridge" . . . . . . . . . . . . . . . . . . . . . . 1 2. From nature to technology . . . . . . . . . . . . . . . . . . . . . 3 3. From technology to nature . . . . . . . . . . . . . . . . . . . . . 4 B. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 II Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 A. Background: Machine learning . . . . . . . . . . . . . . . . . . . . . 10 1. Evolutionary algorithms . . . . . . . . . . . . . . . . . . . . . . . 11 2. Endogenous #tness . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3. Reinforcement learning . . . . . . . . . . . . . . . . . . . . . . . 18 B. Local selection . . . . . . . . . . . . . . . . . . . . . . . . . . ...