Some remarks on the class of continuous (Semi-) strictcly quasiconvex functions

Vegeu el resum a l'inici del document del fitxer adjunt

Generic optimality conditions for semi-algebraic convex programs

We consider linear optimization over a nonempty convex semi-algebraic feasible region F. Semidefinite programming is an example. If F is compact, then for almost every linear objective there is a unique optimal solution, lying on a unique \active" manifold, around which F is \partly smooth", and the second-order sufficient conditions hold. Perturbing the objective results in smooth variation of the optimal solution. The active manifold consists, locally, of these perturbed optimal solutions; it is independent of the representation of F, and is eventually identified by a variety of iterative algorithms such as proximal and projected gradient schemes. These results extend to unbounded sets F.

An algorithm for semi-infinite polynomial optimization

"Vegeu el resum a l'inici del document del fitxer adjunt."

Discrete time Non-homogeneous Semi-Markov Processes applied to Models for Disability Insurance

In this paper, we present a stochastic model for disability insurance contracts. The model is based on a discrete time non-homogeneous semi-Markov process (DTNHSMP) to which the backward recurrence time process is introduced. This permits a more exhaustive study of disability evolution and a more efficient approach to the duration problem. The use of semi-Markov reward processes facilitates the possibility of deriving equations of the prospective and retrospective mathematical reserves. The model is applied to a sample of contracts drawn at random from a mutual insurance company.

Semi-online neural-Q_leaming for real-time robot learning

Reinforcement learning (RL) is a very suitable technique for robot learning, as it can learn in unknown environments and in real-time computation. The main difficulties in adapting classic RL algorithms to robotic systems are the generalization problem and the correct observation of the Markovian state. This paper attempts to solve the generalization problem by proposing the semi-online neural-Q_learning algorithm (SONQL). The algorithm uses the classic Q_learning technique with two modifications. First, a neural network (NN) approximates the Q_function allowing the use of continuous states and actions. Second, a database of the most representative learning samples accelerates and stabilizes the convergence. The term semi-online is referred to the fact that the algorithm uses the current but also past learning samples. However, the algorithm is able to learn in real-time while the robot is interacting with the environment. The paper shows simulated results with the "mountain-car" benchmark and, also, real results with an underwater robot in a target following behavior

Generació semi-automàtica de façanes procedurals

L’objectiu d’aquest PFC és desenvolupar una eina d’edició de façanes procedural apartir d’una imatge d’una façana real. L’aplicació generarà les regles procedurals de lafaçana a partir de dades adquirides del model que es vol representar, com unafotografia. L’usuari de l’aplicació generarà de forma semi-automàtica i interactiva lesregles de subdivisió i repetició, especificant també la inserció de elementsarquitectònics (portes, finestres), que podran ser instanciats a partir d’una llibreria. Uncop generades, les regles s’escriuran en el format del sistema BuildingEngine perintegrar-se completament dins el procés de modelatge urbà.Aquest projecte es desenvoluparà en Matlab

Semi-purity of tempered Deligne cohomology

In this paper we define the formal and tempered Deligne cohomology groups, that are obtained by applying the Deligne complex functor to the complexes of formal differential forms and tempered currents respectively. We then prove the existence of a duality between them, a vanishing theorem for the former and a semipurity property for the latter. The motivation of these results comes from the study of covariant arithmetic Chow groups. The semipurity property of tempered Deligne cohomology implies, in particular, that several definitions of covariant arithmetic Chow groups agree for projective arithmetic varieties.

Impedance field and transition from thermal to shot noise in Cd1-xZnxTe semi-insulating Ohmic detectors

Within a drift-diffusion model we investigate the role of the self-consistent electric field in determining the impedance field of a macroscopic Ohmic (linear) resistor made by a compensated semi-insulating semiconductor at arbitrary values of the applied voltage. The presence of long-range Coulomb correlations is found to be responsible for a reshaping of the spatial profile of the impedance field. This reshaping gives a null contribution to the macroscopic impedance but modifies essentially the transition from thermal to shot noise of a macroscopic linear resistor. Theoretical calculations explain a set of noise experiments carried out in semi-insulating CdZnTe detectors.

Transparent conducting thin films by the co-sputtering of ZnO-ITO targets

Transparent and conductive Zn-In-Sn-O (ZITO) amorphous thin films have been deposited at room temperature by the rf magnetron co-sputtering of ITO and ZnO targets. Co-sputtering gives the possibility to deposit multicomponent oxide thin films with different compositions by varying the power to one of the targets. In order to make ZITO films with different Zn content, a constant rf power of 50 W was used for the ITO target, where as the rf power to ZnO target was varied from 25 W to 150 W. The as deposited films showed an increase in Zn content ratio from 17 to 67 % as the power to ZnO target was increased from 25 to 150 W. The structural, electrical and optical properties of the as deposited films are reported. The films showed an average transmittance over 80% in the visible wavelength range. The electrical resistivity and optical band gap of the ZITO films were found to depend on the Zn content in the film. The ZITO films deposited at room temperature with lower Zn content ratios showed better optical transmission and electrical properties compared to ITO film.

Electrical and optical properties of Zn-In-Sn-O transparent conducting thin films

Indium tin oxide (ITO) is one of the widely used transparent conductive oxides (TCO) for application as transparent electrode in thin film silicon solar cells or thin film transistors owing to its low resistivity and high transparency. Nevertheless, indium is a scarce and expensive element and ITO films require high deposition temperature to achieve good electrical and optical properties. On the other hand, although not competing as ITO, doped Zinc Oxide (ZnO) is a promising and cheaper alternative. Therefore, our strategy has been to deposit ITO and ZnO multicomponent thin films at room temperature by radiofrequency (RF) magnetron co-sputtering in order to achieve TCOs with reduced indium content. Thin films of the quaternary system Zn-In-Sn-O (ZITO) with improved electrical and optical properties have been achieved. The samples were deposited by applying different RF powers to ZnO target while keeping a constant RF power to ITO target. This led to ZITO films with zinc content ratio varying between 0 and 67%. The optical, electrical and morphological properties have been thoroughly studied. The film composition was analysed by X-ray Photoelectron Spectroscopy. The films with 17% zinc content ratio showed the lowest resistivity (6.6 × 10 - 4 Ω cm) and the highest transmittance (above 80% in the visible range). Though X-ray Diffraction studies showed amorphous nature for the films, using High Resolution Transmission Electron Microscopy we found that the microstructure of the films consisted of nanometric crystals embedded in a compact amorphous matrix. The effect of post deposition annealing on the films in both reducing and oxidizing atmospheres were studied. The changes were found to strongly depend on the zinc content ratio in the films.