Integration of prior knowledge of measurement noise in kernel density classification

Samples can be measured with different precisions and reliabilities in different experiments, or even within the same experiment. These varying levels of measurement noise may deteriorate the performance of a pattern recognition system, if not treated with care. Here we seek to investigate the benefit of incorporating prior knowledge about measurement noise into system construction. We propose a kernel density classifier which integrates such prior knowledge. Instead of using an identical kernel for each sample, we transform the prior knowledge into a distinct kernel for each sample. The integration procedure is straightforward and easy to interpret. In addition, we show how to estimate the diverse measurement noise levels in a real world dataset. Compared to the basic methods, the new kernel density classifier can give a significantly better classification performance. As expected, this improvement is more obvious for small sample size datasets and large number of features.     

The Effects of H2O, CO and CO2 on the H2 Permeance and Surface Characteristics of 1 mm Thick Pd80wt%Cu Membranes

The hydrogen permeance of 1 mm-thick Pd_80wt%Cu foils was measured in the presence of equimolar mixtures of H₂ with CO, CO₂ or H₂O over the temperature and total pressure ranges of 623–1,173 K and 0.62–2.86 MPa, respectively. In all cases, permeance losses at 623 and 738 K were very modest. At higher temperatures, more significant decreases in membrane permeance were observed with the highest reduction of about 50% occurring when macroscopic carbon deposition occurred on the membrane surface during H₂–CO exposure at 908 K. The more worrisome effects of exposure to these gases, however, were the micron-scale surface defects observed at 908 and 1,038 K. Although the 1 mm thick disk membranes retained their mechanical integrity, such defects could cause catastrophic failure of ultra-thin, Pd–Cu membranes (1–5 μm thick) deposited on porous substrates.     

Predeployed secure key distribution mechanisms in sensor networks: current state-of-the-art and a new approach using time information

In this article we survey a number of predeployed secure key distribution (PSKD) schemes proposed in the technical literature. We also propose a new time-based PSKD (TPSKD), which operates under the assumption of loose time synchronization, and discuss the performance of the scheme. Since the TPSKD scheme uses time information, which would typically already be available in sensor nodes, the cost of the scheme?s implementation is low.     

Der Schwefelgehalt von Lebensmitteln als analytische Kennzahl

Ohne Zusammenfassung     

Noise–Power Optimization of Incremental Data Converters

Incremental data converters (IDCs) are useful in instrumentation and measurement applications, where low-frequency analog signals need to be converted into digital form with high accuracy and low power dissipation. They are particularly well suited for applications where a single analog-digital converter is multiplexed between many channels. This paper proposes an exact design methodology for IDCs, which optimizes the signal-to-noise ratio of the converter under practical design constraints. The process also allows the designer to apportion the noise budget in an arbitrary manner between thermal and quantization noise. The design process is illustrated by an example which describes the optimization of a third-order multiplexed IDC.     

A Novel Ordered Nanopore Array Diode Laser

In this work, we have created a new type of structure, the nanopore active layer, for achieving quantization of carrier states in a semiconductor. The nanopore structure consists of a periodic two-dimensional array of localized energy barriers perturbing an otherwise conventional quantum well. This perturbation leads to the formation of intraband forbidden energy gaps which are observed experimentally.     

Photo-Fenton removal of water-soluble polymers

This paper presents the results of experiments carried out in a laboratory-scale photochemical reactor on the photodegradation of different polymers in aqueous solutions by the photo-Fenton process. Solutions of three polymers, polyethyleneglicol (PEG), polyacrylamide (PAM), and polyvinylpyrrolidone (PVP), were tested under different conditions. The reaction progress was evaluated by sampling and analyzing the total organic carbon concentration in solution (TOC) along the reaction time. The behavior of the different polymers is discussed, based on the evolution of the TOC–time curves. Under specific reaction conditions, the formation and coalescence of solid particles was visually observed. Solids formation occurred simultaneously to a sharp decrease in the TOC of the liquid phase. This may be favorable for the treatment of industrial wastewater containing polymers, since the photodegradation process can be coupled with solid separation systems, which may reduce the treatment cost.     

Passivity‐based sliding mode control for nonlinear systems

Passivity with sliding mode control for a class of nonlinear systems with and without unknown parameters is considered in this paper. In fact, a method for deriving a nonlinear system with external disturbances to a passive system is considered. Then a passive sliding mode control is designed corresponding to a given storage function. The passivity property guarantees the system stability while sliding mode control techniques assures the robustness of the proposed controller. When the system includes unknown parameters, an appropriate updated law is obtained so that the new transformed system is passive. The passivation property of linear systems with sliding mode is also analysed. The linear and nonlinear theories are applied to a simple pendulum model and the gravity‐flow/pipeline system, respectively. Copyright © 2008 John Wiley & Sons, Ltd.     

A robust algorithm for text region detection in natural scene images

In this paper, a new method for detecting text regions in natural scene images is presented. The proposed algorithm is based on the segmentation of objects in a scene, followed by the identification of text objects by a support vector machine (SVM). First, to segment objects in the scene, the input image is separated into chromatic and achromatic regions according to the distribution of red, green and blue (RGB) elements, and different clustering algorithms are applied. Second, each object is transformed into the wavelet domain for multi-resolution analysis, and moment features of the wavelet coefficients are used in the SVM for the classification of text objects. The proposed approach provides robustness to non-uniform illumination by using different clustering algorithms according to the characteristics of the colour components in the segmentation. Also, moment features, used for classification, are invariant to the size, direction, shape and other properties of texts. Experimental results demonstrate the effectiveness of this approach.