Corrected upper bounds for free-cut elimination

Free-cut elimination allows cut elimination to be carried out in the presence of non-logical axioms. Formulas in a proof are anchored provided they originate in a non-logical axiom or non-logical inference. This paper corrects and strengthens earlier upper bounds on the size of free-cut elimination. The correction requires that the notion of a free cut be modified so that a cut formula is anchored provided that all of its introductions are anchored, instead of only requiring that one of its introductions is anchored. With the correction, the originally proved size upper bounds remain unchanged. These results also apply to partial cut elimination. We also apply these bounds to elimination of cuts in propositional logic.If the non-logical inferences are closed under cut and infer only atomic formulas, then all cuts can be eliminated. This extends earlier results of Takeuti and of Negri and von Plato.     

Evolution strategies with cumulative step length adaptation on the noisy parabolic ridge

This paper presents an analysis of the performance of the (μ/μ,λ)-ES with isotropically distributed mutations and cumulative step length adaptation on the noisy parabolic ridge. Several forms of dependency of the noise strength on the distance from the ridge axis are considered. Closed form expressions are derived that describe the mutation strength and the progress rate of the strategy in high-dimensional search spaces. It is seen that as for the sphere model, larger levels of noise present lead to cumulative step length adaptation generating increasingly inadequate mutation strengths, and that the problem can be ameliorated to some degree by working with larger populations.     

Exploring the Interaction between Java's Implicitly Thrown Exceptions and Instruction Scheduling

The frequent occurrence of implicitly thrown exceptions poses one of the challenges present in a Java compiler. Not only do these implicitly thrown exceptions directly affect the performance by requiring explicit checks, they also indirectly impact the performance by restricting code movement in order to satisfy the precise exception model in Java. In particular, instruction scheduling is one transformation that is restricted by implicitly thrown exceptions due to the heavy reliance on reordering instructions to exploit maximum hardware performance. The goal of this study is two-fold: first, investigate the degree to which implicitly thrown exceptions in Java hinder instruction scheduling, and second, find new techniques for allowing more efficient execution of Java programs containing implicitly thrown exceptions. Experimental results show that with aggressive scheduling techniques, such as superblock scheduling, the negative performance impact can be greatly reduced.     

Optimum tracking with evolution strategies

Evolutionary algorithms are frequently applied to dynamic optimization problems in which the objective varies with time. It is desirable to gain an improved understanding of the influence of different genetic operators and of the parameters of a strategy on its tracking performance. An approach that has proven useful in the past is to mathematically analyze the strategy's behavior in simple, idealized environments. The present paper investigates the performance of a multiparent evolution strategy that employs cumulative step length adaptation for an optimization task in which the target moves linearly with uniform speed. Scaling laws that quite accurately describe the behavior of the strategy and that greatly contribute to its understanding are derived. It is shown that in contrast to previously obtained results for a randomly moving target, cumulative step length adaptation fails to achieve optimal step lengths if the target moves in a linear fashion. Implications for the choice of population size parameters are discussed.     

Quasiparticle tunneling through rare earth trihydroxide barriers

Symmetric tunneling junctions with 4000-Å-thick Pb electrodes and polycrystalline insulating barriers of Lu(OH)₃, Er(OH)₃, and Ho(OH)₃ have been fabricated. In bulk, these three rare earth trihydroxides are nonmagnetic, antiferromagnetic (T _N<1.1 K), and ferromagnetic (T_c=2.54 K), respectively. Tunneling resistances ranged from 600 to greater than 40,000 with a junction area of 6.25×10^–2 cm². Single-particle tunneling characteristics of these junctions were always broadened relative to the characteristics of Pb-PbO-Pb junctions, although the ratio of the zero-bias tunneling resistance to the normal tunneling resistance in some instances was of the order of 1000. A threefold splitting of the conductance peak at the gap was observed only in junctions with Ho(OH)₃ barriers. The gap peak of junctions with Er(OH)₃ barriers was broadened significantly relative to that of junctions with Lu(OH)₃ barriers. From measurements of the temperature and magnetic field dependences of the tunneling conductance it is argued that the splitting in junctions with Ho(OH)₃ barriers is consistent with the existence of a peak in the electronic density of states at an energy below that of the gap of each of the electrodes. This peak is believed to be the signature of a bound state near the barrier where the pair potential is depressed by virtue of the exchange coupling between the spins of the superconducting electrons and the localized spins of the barrier. Qualitative interpretations of the data support the view that the observed structure in Ho(OH)₃ barrier junctions is neither a consequence of intrinsic gap anisotropy in Pb nor of inelastic magnon-assisted tunnelling.     

Data encodings and their costs

Summary This paper is devoted to developing and studying a precise notion of the encoding of a logical data structure in a physical storage structure, that is motivated by considerations of computational efficiency. The development builds upon the notion of an encoding of one graph in another. The cost of such an encoding is then defined so as to reflect the structural compatibility of the two graphs, the (externally specified) costs of implementing the host graph, and the (externally specified) set of intended usage patterns of the guest graph. The stability of the constructed framework is demonstrated in terms of a number of results; the faithfulness of the formalism is argued in terms of a number of examples from the literature; and the tractability of the model is hinted at by several results and by further references to the literature.     

A mobility constraint model to infer sensor behaviour in forest fire risk monitoring

Wireless sensor networks (WSNs) play an important role in forest fire risk monitoring. Various applications are in operation. However, the use of mobile sensors in forest risk monitoring remains largely unexplored. Our research contributes to fill this gap by designing a model which abstracts mobility constraints within different types of contexts for the inference of mobile sensor behaviour. This behaviour is focused on achieving a suitable spatial coverage of the WSN when monitoring forest fire risk. The proposed mobility constraint model makes use of a Bayesian network approach and consists of three components: (1) a context typology describing different contexts in which a WSN monitors a dynamic phenomenon; (2) a context graph encoding probabilistic dependencies among variables of interest; and (3) contextual rules encoding expert knowledge and application requirements needed for the inference of sensor behaviour. As an illustration, the model is used to simulate the behaviour of a mobile WSN to obtain a suitable spatial coverage in low and high fire risk scenarios. It is shown that the implemented Bayesian network within the mobility constraint model can successfully infer behaviour such as sleeping sensors, moving sensors, or deploying more sensors to enhance spatial coverage. Furthermore, the mobility constraint model contributes towards mobile sensing in which the mobile sensor behaviour is driven by constraints on the state of the phenomenon and the sensing system.     

Implementing molecular dynamics on hybrid high performance computers – Particle–particle particle-mesh

The use of accelerators such as graphics processing units (GPUs) has become popular in scientific computing applications due to their low cost, impressive floating-point capabilities, high memory bandwidth, and low electrical power requirements. Hybrid high-performance computers, machines with nodes containing more than one type of floating-point processor (e.g. CPU and GPU), are now becoming more prevalent due to these advantages. In this paper, we present a continuation of previous work implementing algorithms for using accelerators into the LAMMPS molecular dynamics software for distributed memory parallel hybrid machines. In our previous work, we focused on acceleration for short-range models with an approach intended to harness the processing power of both the accelerator and (multi-core) CPUs. To augment the existing implementations, we present an efficient implementation of long-range electrostatic force calculation for molecular dynamics. Specifically, we present an implementation of the particle–particle particle-mesh method based on the work by Harvey and De Fabritiis. We present benchmark results on the Keeneland InfiniBand GPU cluster. We provide a performance comparison of the same kernels compiled with both CUDA and OpenCL. We discuss limitations to parallel efficiency and future directions for improving performance on hybrid or heterogeneous computers.     

Drinking patterns as predictors of alcohol withdrawal reactions in DBA/2J mice

Individually housed DBA/2J mice were fed a liquid diet in which ethanol supplied 33% of the calories. The level of physical dependence that developed was estimated by scoring convulsions, elicited by handling the mice, after discontinuing the alcohol diet. The severity of the withdrawal reaction increased progressively with duration (5-12 days) of alcohol administration. A 2-day period on the diet produced no withdrawal reaction. Pretreatment of the mice with alcohol in their drinking water slightly increased the subsequent intake of the liquid diet. "Effective" alcohol intake was defined as uninterrupted alcohol consumption above 10 g/kg/day. Withdrawal scores correlated better with effective intake than with total intake under a variety of conditions. We interpret this to mean that brief interruptions in drinking (1 day) may allow the accrued physical dependence to disappear. On the basis of their effective alcohol intake, mice could be assigned to nondependent, moderately dependent or severely dependent groups for further study of the nature of physical dependence.