Search

In this paper we explore the Vs-optimality criterion that was proposed in Ginsburg and Ben-Gal (IIE Trans. 2006; 38:445–461) as a new design-of-experiment (DOE) alphabetic optimality criterion. The Vs-optimality criterion seeks to minimize the variance of the optimal solution of an empirically fitted model. We show that the Vs-optimality citerion is well related to known alphabetic DOE criteria. However, it focuses on an 'optimization-seeking' experimental approach rather than an 'information-seeking' approach, which is often adopted by traditional optimality criteria. We illustrate the differences between these two approaches by a detailed example of a robust design of an RL-circuit. Copyright © 2009 John Wiley & Sons, Ltd. [ABSTRACT FROM AUTHOR]

This paper points out that meta-heuristics should have not only robustness and adaptability to problems with different structure but also adjustability of parameters included in their algorithms. Particle swarm optimization (PSO), whose concept began as a simulation of a simplified social milieu, is known as one of the most powerful optimization methods for solving nonconvex continuous optimization problems. Then, in order to improve adjustability, a new parameter is introduced into PSO on the basis of the proximate optimality principle (POP). In this paper, we propose adaptive PSO and the effectiveness and the feasibility of the proposed approach are demonstrated on simulations using some typical nonconvex optimization problems. © 2005 Wiley Periodicals, Inc. Electr Eng Jpn, 151(3): 41–49, 2005; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/eej.20077 [ABSTRACT FROM AUTHOR]

The paper presents an analytical description of a state-dependent assignment policy, where different assignments of attributes to objects are employed under different states. Various considerations, other than optimizing an objective function, present themselves for such cases—for example, the existence of groups of assignees with certain levels of intragroup assignment uniformity, the stability of the policy over states, the presence of a revealed preference ordering of the assignments, and others. An approach towards a quantitative treatment of such features of a state-dependent assignment policy is discussed, and in particular a uniformity measure is defined and its properties are investigated. Potential uses of the model are suggested for urban emergency services, questionnaire analysis, task allocation, medical diagnosis, and dynamic computer program behavior. Three implementations for questionnaire analysis are reported. [ABSTRACT FROM AUTHOR]

Statistical inference is a branch of mathematics, but is not an intrinsic part of substantive science. Rather, it belongs to the scaffolding. Although statistical inference is based on probabilistic notions, the probabilities used in such inference might be termed artificial in contrast to the natural probabilities of substantive phenomena. Statistical inference is concerned with sampling distributions, that is, with what would occur if infinitely many random samples were drawn from the same population.