A Comparison of the Rough Sets and Recursive Partitioning Induction Approaches: An Application to Commercial Loans

Credit scoring is the term used to describe methods utilized for classifying applicants for credit into classes of risk. This paper evaluates two induction approaches, rough sets and decision trees, as techniques for classifying credit (business) applicants. Inductive learning methods, like rough sets and decision trees, have a better knowledge representational structure than neural networks or statistical procedures because they can be used to derive production rules. If decision trees have already been used for credit granting, the rough sets approach is rarely utilized in this domain. In this paper, we use production rules obtained on a sample of 1102 business loans in order to compare the classification abilities of the two techniques. We show that decision trees obtain better results with 87.5% of good classifications with a pruned tree, against 76.7% for rough sets. However, decision trees make more type–II errors than rough sets, but fewer type–I errors.     

Data fitting with geometric-programming-compatible softmax functions

Motivated by practical applications in engineering, this article considers the problem of approximating a set of data with a function that is compatible with geometric programming (GP). Starting with well-established methods for fitting max-affine functions, it is shown that improved fits can be obtained using an extended function class based on the softmax of a set of affine functions. The softmax is generalized in two steps, with the most expressive function class using an implicit representation that allows fitting algorithms to locally tune softness. Each of the proposed function classes is directly compatible with the posynomial constraint forms in GP. Max-monomial fitting and posynomial fitting are shown to correspond to fitting special cases of the proposed implicit softmax function class. The fitting problem is formulated as a nonlinear least squares regression, solved locally using a Levenberg–Marquardt algorithm. Practical implementation considerations are discussed. The article concludes with numerical examples from aerospace engineering and electrical engineering.     

The MORPHEUS Heterogeneous Dynamically Reconfigurable Platform

Reconfigurable computing offers a wide range of low cost and efficient solutions for embedded systems. The proper choice of the reconfigurable device, the granularity of its processing elements and its memory architecture highly depend on the type of application and their data flow. Existing solutions either offer fine grain FPGAs, which rely on a hardware synthesis flow and offer the maximum degree of flexibility, or coarser grain solutions, which are usually more suitable for a particular type of data flow and applications. In this paper, we present the MORPHEUS architecture, a versatile reconfigurable heterogeneous System-on-Chip targeting streaming applications. The presented architecture exploits different reconfigurable technologies at several computation granularities that efficiently address the different applications needs. In order to efficiently exploit the presented architecture, we implemented a complete software solution to map C applications to the reconfigurable architecture. In this paper, we describe the complete toolset and provide concrete use cases of the architecture.     

Implications of short and long term (30 days) sorption on the desorption kinetic of trace metals (Cd, Zn, Co, Mn, Fe, Ag, Cs) associated with river suspended matter

The desorption kinetic of trace elements (Cd, Zn, Co, Mn, Fe, Ag, and Cs) associated with Loire river natural suspended particulate matter (SPM; 0.4-63 microm) was followed up on times varying from 0.5 h to 30 days, from SPM previously contaminated during 1 h, 24 h and 30 days. Long term sorption kinetics indicated that the difference between sorption occurring during the period 0-30 days (time investigated in this study) and the period 0-48 h (time often used for sorption experiments) ranges from few to 25% according to the element. Desorption kinetics show that, whatever the age of the complex formed during the sorption step, the release tends to equilibrium between complexed and dissolved elements equivalent to the equilibrium obtained for sorption after a given time. However, the time to get this equilibrium depends on the aging of the complex and on the element. All the above features indicate different types of complexes formation and strength of the binding according to the age of the complex and according to the element. Using a multi-compartmental model, simulating the transfer of metals between water and different types of particulate sites, the relationships between the parameters describing slow and rapid processes helped in explaining the "aging" effect observed.     

Diffraction theory of an anisotropic circular cylinder

A fully analytical theory is developed to derive the field diffracted by an infinitely long circular cylinder made of an arbitrary anisotropic homogeneous material, illuminated by an arbitrary plane wave.     

Shannon entropy of partially polarized and partially coherent light with Gaussian fluctuations

We propose to analyze Shannon entropy properties of partially coherent and partially polarized light with Gaussian probability distributions. It is shown that the Shannon entropy is a sum of simple functions of the intensity, of the degrees of polarization, and of the intrinsic degrees of coherence that have been recently introduced. This analysis clearly demonstrates the contribution of partial polarization and of partial coherence to the characterization of disorder of the light provided by the Shannon entropy, which is a standard measure of randomness. We illustrate these results on two simple examples.     

Hierarchically structured transparent hybrid membranes by in situ growth of mesostructured organosilica in host polymer

The elaborate performances characterizing natural materials result from functional hierarchical constructions at scales ranging from nanometres to millimetres, each construction allowing the material to fit the physical or chemical demands occurring at these different levels. Hierarchically structured materials start to demonstrate a high input in numerous promising applied domains such as sensors, catalysis, optics, fuel cells, smart biologic and cosmetic vectors. In particular, hierarchical hybrid materials permit the accommodation of a maximum of elementary functions in a small volume, thereby optimizing complementary possibilities and properties between inorganic and organic components. The reported strategies combine sol-gel chemistry, self-assembly routes using templates that tune the material's architecture and texture with the use of larger inorganic, organic or biological templates such as latex, organogelator-derived fibres, nanolithographic techniques or controlled phase separation. We propose an approach to forming transparent hierarchical hybrid functionalized membranes using in situ generation of mesostructured hybrid phases inside a non-porogenic hydrophobic polymeric host matrix. We demonstrate that the control of the multiple affinities existing between organic and inorganic components allows us to design the length-scale partitioning of hybrid nanomaterials with tuned functionalities and desirable size organization from ?ngstr?m to centimetre. After functionalization of the mesoporous hybrid silica component, the resulting membranes have good ionic conductivity offering interesting perspectives for the design of solid electrolytes, fuel cells and other ion-transport microdevices.     

Self-assembly and characterization of layered double hydroxide/DNA hybrids

The purpose of this study was to control the fabrication of new labile supramolecular assemblies by formulating associations of DNA molecules with inorganic layered double hydroxides (LDHs). The results show that LDH/DNA hybrids synthesized by a coprecipitation route involving the in situ formation of LDHs around DNA molecules acting as templates were characterized by a lamellar organization, with DNA molecules sandwiched between hydroxide layers, exhibiting a regular spacing of 1.96 nm. Our results indicate that labile complexes resulting from the association of nucleic acids and inorganic materials can be obtained not only by anion exchange but also by a direct self-assembly route.     

The complexity of mean flow time scheduling problems with release times

We study the problem of preemptive scheduling of n} jobs with given release times on m identical parallel machines. The objective is to minimize the average flow time. In this paper, show that when all jobs have equal processing times then the problem can be solved in polynomial time using linear programming. Our algorithm can also be applied to the open-shop problem with release times and unit processing times. For the general case (when processing times are arbitrary), we show that the problem is unary NP-hard. P. Baptiste and C. Dürr: Supported by the NSF/CNRS grant 17171 and ANR/Alpage. P. Brucker: Supported by INTAS Project 00-217 and by DAAD PROCOPE Project D/0427360. M. Chrobak: Supported by NSF grants CCR-0208856 and INT-0340752. S. A. Kravchenko: Supported by the Alexander von Humboldt Foundation.