Search

International audience; Genetic Regulatory Networks (GRNs) implementations have a high degree of variability in their details. Parameters, encoding methods, and dynamics formulas all differ in the literature, and some GRN implementations have a high degree of model complexity. In this paper, we present a comparative study of different implementations of a GRN and introduce new variants for comparison. We use a modified Genetic Algorithm (GA) to evaluate GRN performance on a number of common benchmark tasks, with a focus on real-time control problems. We propose an encoding scheme and set of dynamics equations that simplifies implementation and evaluate the evolutionary fitness of this proposed method. Lastly, we use the comparative modifications study to demonstrate overall enhancements for GRN models.

Using a fast numerical technique, we investigate a large database of investors’ suboptimal nonexercise of short-maturity American call options on dividend-paying stocks listed on the Dow Jones. The correct modeling of the discrete dividend is essential for a correct calculation of the early exercise boundary, as confirmed by theoretical insights. Pricing with stochastic volatility and jumps instead of the Black–Scholes–Merton benchmark cuts the amount lost by investors through suboptimal exercise by one-quarter. The remaining three-quarters are largely unexplained by transaction fees and may be interpreted as an opportunity cost for the investors to monitor optimal exercise.

Usually, the blood cell counting activity in haematology laboratory uses the comparison of IQC values to the target values proposed by the manufacturer. We intended to improve the monitoring of the proper functioning of our analytical measure system for 17 main haematologic parameters. To set the allowable critical limits of IQC, we propose our reflection based on several elements: benchmark and expert recommendation, clinical requirements, statistical indicators of the laboratory calculated using IQC values (3 levels, 2 different lots, 2 haematology analysers and 2 passage modes) and the EEQ values, during four months. We exploited the reports obtained from the middleware (our own IQC values), and the interlaboratory comparison reports (obtained from SNCS and EuroCell websites) and we compared our performances to the Ricos objectives, to set clearly argued allowable limits for IQC values. Finally, the allowable limits correspond to the imprecision limits stated by Ricos for 14 parameters (desirable for 11 parametres and minimal for 3 parameters) and personalized limits (more exigent than desirable Ricos limits) for 3 parameters of blood cell counting.

Contrary to the classical continuous media of metallic structures, fibrous composites have a very specific mechanical behaviour due to their composition. The prediction of the properties in simulating the forming processes of woven reinforcements necessitates special analysis methods. The objective of this research work is to present the continuous approach which can be exploited within a commercial code (e.g. ABAQUS, used here). For that we treat, primarily, the elementary tests with continuous approach in large deformations with three different methods i.e. single element with unidirectional fibers, single element with bidirectional fibers and two superimposed elements with unidirectional fibers. The tests are performed using locally developed user material subroutine (VUMAT) for membrane elements in ABAQUS/Explicit. The numerical results of the elementary tests conform with each other and to the exact analytical solutions. Also, these tests are extended to the bias extension test, picture frame test, hemispherical dome forming and the international benchmark draping of double dome. The numerical outputs reasonably corroborate with experimental tests.

Smart card is today the most widespread secured portable computing device. Four years ago, we addressed the problem of scaling down database techniques for the smart card and we proposed the design of what we called a PicoDBMS, a full-fledged database system embedded in a smart card. Since then, thanks to the hardware progress and to the joint implementation efforts of our team and our industrial partner, this utopian design gave birth to a complete prototype running on an experimental smart card platform. This paper revisits the problem statement in the light of the hardware and applications evolution. Then, it introduces a benchmark dedicated to Pico-style databases and provides an extensive performance analysis of our prototype, discussing lessons learned at experimentation time and helping selecting the appropriate storage and indexation model for a given class of embedded applications. Finally, it draws new research perspectives for data management on secured chips. MOTS-CLES : Bases de donnees embarquees, Confidentialite et protection des donnees, Carte a puce, Controle d'acces, Modeles de stockage et d'indexation, Evaluation de requetes.

International audience; Information visualization needs benchmarks to carry on. A benchmark is aimed at comparing information visualization techniques or systems. A benchmark is made of a dataset, a list of tasks mostly based on finding facts about the dataset, and a list of interesting or important findings about the datasets (the nuggets to find). For the second year, we are organizing the InfoVis Contest aimed at collecting results for benchmarks. We describe here the main lessons we learned.