Probabilistic consensus clustering using evidence accumulation

Machine Learning - Clustering ensemble methods produce a consensus partition of a set of data points by combining the results of a collection of base clustering algorithms. In the evidence...     

GaN‐based emissive microdisplays: A very promising technology for compact,ultra‐high brightness display systems

High‐brightness GaN‐based emissive microdisplays can be fabricated with different approaches that are listed and described. They consist either of hybridizing a GaN LED array on a CMOS circuit or building a monolithic component on a single substrate. Using the hybridization approach, two types of 10‐μm pixel pitch GaN microdisplay prototypes were developed: (1) directly driven, 300 × 252 pixels and (2) active‐matrix, 873 × 500 pixels. Brightness as high as 1 × 10⁶ and 1 × 10⁷ cd/m² for blue and green arrays, respectively, were reached. GaN‐based emissive microdisplays are suitable for augmented reality systems or head‐up displays, but some challenges remain before they can be put in production.     

A projected gradient method for optimization over density matrices

An ensemble of quantum states can be described by a Hermitian, positive semidefinite and unit trace matrix called density matrix. Thus, the study of methods for optimizing a certain function (energy, entropy) over the set of density matrices has a direct application to important problems in quantum information and computation. We propose a projected gradient method for solving such problems. By exploiting the geometry of the feasible set, which is the intersection of the cone of Hermitian positive semidefinite matrices with the hyperplane defined by the unit trace constraint, we describe an efficient procedure to compute the projection onto this set using the Frobenius norm. Some important applications, such as quantum state tomography, are described and numerical experiments illustrate the effectiveness of the method when compared to previous methods based on fixed-point iterations or semidefinite programming.     

An adaptive support vector regressor controller for nonlinear systems

In this study, a novel online support vector regressor (SVR) controller based on system model estimated by a separate online SVR is proposed. The main idea is to obtain an SVR controller based on an estimated model of the system by optimizing the margin between reference input and system output. For this purpose, “closed-loop margin” which depends on tracking error is defined, then the parameters of the SVR controller are optimized so as to optimize the closed-loop margin and minimize the tracking error. In order to construct the closed-loop margin, the model of the system estimated by an online SVR is utilized. The parameters of the SVR controller are adjusted via the SVR model of system. The stability of the closed-loop system has also been analyzed. The performance of the proposed method has been evaluated by simulations carried out on a continuously stirred tank reactor (CSTR) and a bioreactor, and the results show that SVR model and SVR controller attain good modeling and control performances.     

Reordering space design in statistical machine translation

In Statistical Machine Translation (SMT), the constraints on word reorderings have a great impact on the set of potential translations that is explored during search. Notwithstanding computational issues, the reordering space of a SMT system needs to be designed with great care: if a larger search space is likely to yield better translations, it may also lead to more decoding errors, because of the added ambiguity and the interaction with the pruning strategy. In this paper, we study the reordering search space, using a state-of-the art translation system, where all reorderings are represented in a permutation lattice prior to decoding. This allows us to directly explore and compare different reordering schemes and oracle settings. We also study in detail a rule-based preordering system, varying the length and number of rules, the tagset used, as well as contrasting with purely combinatorial subsets of permutations. We carry out experiments on three language pairs in both directions: English-French, a close language pair; English-German and English-Czech, two much more challenging pairs. We show that even though it might be desirable to design better reordering spaces, model and search errors seem to be the most important issues. Therefore, improvements of the reordering space should come along with improvements of the associated models to be really effective.     

Terrain Modelling from Feature Primitives

We introduce a compact hierarchical procedural model that combines feature‐based primitives to describe complex terrains with varying level of detail. Our model is inspired by skeletal implicit surfaces and defines the terrain elevation function by using a construction tree. Leaves represent terrain features and they are generic parametrized skeletal primitives, such as mountains, ridges, valleys, rivers, lakes or roads. Inner nodes combine the leaves and subtrees by carving, blending or warping operators. The elevation of the terrain at a given point is evaluated by traversing the tree and by combining the contributions of the primitives. The definition of the tree leaves and operators guarantees that the resulting elevation function is Lipschitz, which speeds up the sphere tracing used to render the terrain. Our model is compact and allows for the creation of large terrains with a high level o detail using a reduced set of primitives. We show the creation of different kinds of landscapes and demonstrate that our model allows to efficiently control the shape and distribution of landform features.     

Performance of an Adaptive Controller for the Neuromuscular Blockade Based on Inversion of a Wiener Model

An adaptive controller based on a minimally parameterized parsimonious Wiener model for the effect of the muscle relaxant rocuronium in the neuromuscular blockade is presented. The controller structure combines inversion of the recursively identified static nonlinearity of the Wiener model with a positive compartmental control law for the linearized system. The overall strategy exploits the fact that the model has only two parameters, which are estimated by an extended Kalman filter. Due to the fact that the positive control law for total mass conservation of compartmental systems is only proven to be convergent for time‐invariant systems, the identification of the parameter in the linear block of the minimally parameterized parsimonious Wiener model is stopped when the controller is turned on. The controller was implemented in the platform Galeno and tested in simulation and in thirteen real cases of patients under general anesthesia. The good reference tracking results and robustness to noise show the reliability of the proposed strategy.     

The magnetic-distortion probe: velocimetry in conducting fluids

A new type of velocimeter, capable of local velocity measurements in conducting fluids, is introduced. The principle of the "magnetic-distortion probe" is based on the measurement of the induced magnetic field by the flow of a conducting fluid in the vicinity of a localized magnetic field. The new velocimeter has no moving parts, and can be enclosed in a sealed cap, easing the implementation in harsh environments, such as liquid metals. The proposed method allows one to probe both the continuous part and fluctuations of the velocity, the temporal and spatial resolution being linked to the actual geometric configuration of the probe. A prototype probe has been tested in a gallinstan pipe flow and in a fully turbulent flow of liquid gallium generated by the counter rotation of two coaxial impellers in a cylinder. The signals have been compared to a reference potential probe and show very good agreement both for time-averaged velocities and turbulent fluctuations. The prototype is shown to detect motion from a few cm s(-1) to a few m s(-1). Moreover, the use of the magnetic-distortion probe with large-scale applied magnetic field is discussed.