Catálogo de publicaciones - libros

Optimizing product lines is the task of planning the joint offer of multiple substitutes concurrently. Actually, this resembles the design of a choice menu where consumers are supposed to choose at most one item from a set of alternatives.

In this paper we model the product line optimization problem by decomposing it into two stages. Product decision is done in the first stage by anticipating possible outcomes of the subsequent (price) stage. This so-called hierarchical decision situation is considered due to uncertainty about consumers’ tastes. Results suggest that substantial increase of expected profit can be drawn from the hierarchical approach.

Within this paper, the energy model PERSEUS-HYDRO and its application to Switzerland with a focus on the possibility to provide long-term electricity prices will be presented. Within this model perfect energy markets are assumed and thus electricity prices can be derived from system-marginal-costs. Main features of the plant dispatch within a system with hydro and thermal power will be shown. Also, the integration of energy exchanges (e.g. EEX) are foreseen within the optimising model. As the plant dispatch mainly influences the marginal costs, the important effects and coherences will be highlighted. As prices on the energy exchange have to be set as exogenous input, their influence on the marginal costs will also be discussed.

A major food safety issue in pork is Salmonella contamination. A stochastic state-transition simulation model was described to simulate the spread of Salmonella from multiplying through slaughter, with special emphasis for critical control points to prevent or reduce Salmonella contamination. Design of Experiments and metamodelling were used for a sensitivity analysis. The finishing stage and the slaughterhouse appeared to be the most important stages in the supply chain to reduce the prevalence of Salmonella contaminated carcasses.

Biological neural system can be considered as a series of biochemical reactions and signal transmission. It is important to provide an intuitive representation of the neural system to biologists while keeping its computational consistency. In this paper, we propose a method to exploit Hybrid Petri Net (HPN) for intuitive representation and quantitative modeling. The HPN is an extension of Petri Nets and represented by a directed, bipartite graph in which nodes are either discrete/continuous places (such as ion channels) or discrete/continuous transitions (such as phosphorylation), where places represent conditions and transitions represent activities. It can easily model the interactions among receptors, ionic flows (such as calcium), G-proteins, protein kinases and transcription factors that are very complicate in terms of the dynamics of all participants and their correlations. We demonstrate that, in the biological neural system, it is possible to translate and map these complex phenomena into HPNs in a natural manner. In our model, the dynamic properties of the neural signal processing can be examined, especially the interactions among neural modulators and signal transduction pathways. With such a mechanism model in hand, our ability to collaborate with neural scientists is greatly enhanced so as to simulate and examine the robustness of the neural transmission under the local biochemical perturbations.

This study concerns modeling, approximation and inference of gene regulatory dynamics on the basis of gene expression patterns. The dynamical behavior of gene expressions is represented by a system of ordinary differential equations. We introduce a gene-interaction matrix with some nonlinear entries, in particular, quadratic polynomials of the expression levels to keep the system solvable. The model parameters are determined by using optimization. Then, we provide the time-discrete approximation of our time-continuous model. Finally, from the considered models we derive gene regulatory networks, discuss their qualitative features and provide a basis for analyzing networks with nonlinear connections.

In this study the joint dynamics between stock prices and trading volume are investigated using data from the German stock market. Our results indicate no relations (contemporaneous as well as dynamic) between return levels and trading volume but strong linkages between return volatility and volume data. On including trading volume in the conditional volatility framework (GARCH-type) we provide empirical evidence for the importance of volume data as an indicator for the flow of information on the market. Applying Granger’s test for causality we detect also feedback relations between trading volume and return volatility. These findings corroborate our assumption that trading volume indirectly contains information about stock prices due to its relation to return volatility.

The generating function of the autocorrelations of successive waiting times in a stationary M/G/l or in a stationary GI/M/1 system can be expressed in terms of the probability generating function of the number of customers served in a busy period. The latter function is only implicitly determined as a solution to a functional equation. More explicit expressions have been obtained with the aid of Lagrange’s theorem on the reversion of power series, but they involve increasingly higher order derivatives of a function which comprises several Laplace-Stieltjes transforms. A recently discovered substitution method for contour integrals allows the numerical inversion of an implicitly determined generating function without the numerical solution of the functional equation for many complex values.

We consider multiobjective optimization problems in which objective functions are in the form of mathematical expectation of functions depending on a random element and a constraints set can depend on a probability measure. An efficient points set characterizes the multiobjective problems very often instead of the solution set in one objective case. A stability of the efficient points set (w.r.t. a probability measures space) and empirical estimates have been already investigated in the case when all objective functions were assumed to be strongly convex. The aim of the contribution is to present a modified assertions under rather weaker assumptions.

Consider the following situation of two-step shortlisting: two experts Alice and Bob are faced with a large number of alternatives which they can only observe imprecisely. They have to choose one of the alternatives, without knowing which one is best. Alice first compiles a shortlist of alternatives by choosing her best observations. Bob then chooses his best observation among the shortlisted alternatives. Previous research showed that this procedure sometimes yielded worse results than if a single expert made the entire decision himself. Here, we consider an urn containing — 1 homogeneous balls and one ball with larger weight. When drawing balls at random from the urn, the probability of drawing any one ball is proportional to its weight. Alice draws balls and puts them in another urn, from which Bob then draws a single ball. Which value of maximizes the probability that Bob draws the distinguished ball?

This note studies the variance of total cumulative rewards for Markov reward chains in both discrete and continuous time. It is shown that parallel results can be obtained for both cases.

First, explicit formulae are presented for the variance within finite time. Next, the infinite time horizon is considered. Most notably, it is concluded that the variance has a linear growth rate. Explicit expressions are provided, related to the standard average reward case, to compute this growth rate.