Cluster-based instance selection for machine classification

Instance selection in the supervised machine learning, often referred to as the data reduction, aims at deciding which instances from the training set should be retained for further use during the learning process. Instance selection can result in increased capabilities and generalization properties of the learning model, shorter time of the learning process, or it can help in scaling up to large data sources. The paper proposes a cluster-based instance selection approach with the learning process executed by the team of agents and discusses its four variants. The basic assumption is that instance selection is carried out after the training data have been grouped into clusters. To validate the proposed approach and to investigate the influence of the clustering method used on the quality of the classification, the computational experiment has been carried out.     

Carbon Monoxide Oxidation over Au/Ce1−x Zr x O2 Catalysts: Effects of Moisture Content in the Reactant Gas and Catalyst Pretreatment

The Au/Ce_1?x Zr _x O₂ (x = 0, 0.25, 1) catalysts were synthesized, characterized by BET, XRD, TPR-H₂, HRTEM, AAS and tested in CO oxidation. The effect of moisture in the reactant gas on CO conversion has been studied in a wide range of concentrations (~0.7–6000 ppm). Moisture generates a positive effect on catalytic activity and wet conditions gave higher CO conversions. The optimum concentration of moisture for CO oxidation over Au/CeO₂ and Au/Ce_0.75Zr_0.25O₂ is 200–1000 ppm, while further increase in the moisture content suppresses CO conversion. The activity of the studied Au catalysts depends on the amount of moisture adsorbed on the catalyst rather than on its content in the feed stream, which suggests that the reaction involves water-derived species on the catalysts surface. The effect of the catalysts pretreatment in air, dry He, H₂ stream as well as H₂ + H₂O gas mixture on their catalytic performance in CO oxidation has been also investigated. The model of the active sites for CO oxidation over the studied catalysts was proposed.     

Multilayer Organic Multichip Module Implementing Hybrid Microelectromechanical Systems

We present the design and development of an organic package that is compatible with fully released RF microelectromechanical systems (MEMS). The multilayer organic package consists of a liquid-crystal polymer film to provide near hermetic cavities for MEMS. The stack is further built up using organic thin-film polyimide. To demonstrate the organic package, we have designed and implemented a 2-bit true-time delay X-band phase shifter using commercially available microelectromechanical switches. The packaged phase shifter has a measured insertion loss of 2.45 plusmn 0.12 dB/bit at 10 GHz. The worst case phase variation of the phase shifter at 10 GHz is measured to less than 5deg. We have also conducted temperature cycling (-65degC to 150degC) and 85/85 to qualify the packaging structures.     

Particle motion and heat transfer in a pulsating gas stream

Results are given of an analytical investigation of the motion of granular material and interphase heat transfer in an ascending two-phase stream with sinusoidal gas velocity fluctuations.     

Study of the morphological and magnetic microstructure of SmCo5 magnets

In this paper, high-coercivity sintered SmCo₅ permanent magnets were investigated. The study of the morphological microstructure was performed by scanning electron microscopy (SEM) using backscattered electron imaging and energy dispersive X-ray (EDX) microanalysis, and the magnetic microstructure was observed with magnetic force microscopy (MFM) in the thermally demagnetized state of the magnets at the surface perpendicular to the alignment axis. In addition to the main SmCo₅ phase, an appreciable amount of a mixture of the Sm₂Co₇ and Sm₅Co₁₉ phases, pores, Sm oxides and carbon was detected. The grains exhibited magnetic domains. Except for rare cases, the domains were continuous from grain to grain, indicating good magnetic alignment of the individual grains.     

Heat and mass transfer during drying of a bed of shrinking particles – Simulation for carrot cubes dried in a spout-fluidized-bed drier

The purpose of the present study was to develop a model to describe the heat and mass transfer during the drying of carrot cubes in a spout-fluidized-bed drier. The model took into account the non-homogeneous shrinkage of the material. The Arbitrary Lagrange–Eulerian (ALE) formulation was applied to enter the problem with moving boundaries. Three phases of drying were distinguished according to the behavior of changes in percent local error of estimation: an initial phase of warming up the material – characterized by a low level of error of moisture content prediction, a second phase – characterized by an increase in the error of moisture content prediction and a phase of decreasing error. A simple test of the sensitivity of the model to the changes in heat transfer coefficient was performed in order to improve the ability of the model to predict the changes in moisture content and temperature of dried carrots. The predicted changes in both the moisture content and the temperature of carrot cubes during drying in a spout-fluidized-bed drier indicate that the model can be successfully applied to describe moisture content, temperature and deformation of dried particles in cases when the very high accuracy of moisture content and temperature prediction is not a crucial element of investigation of the drying process.     

Approach to calculation time-dependent moisture diffusivity for thin layered biological materials

A method is presented to determine the effective diffusivity from experimental drying kinetics as a time-dependent parameter. The method combines an analytical solution of Fick's equation in which the Fourier number is approximated using the empirical coefficients a and b with a semi-theoretical equation derived for quasi-stationary conditions. The resulting equation has been applied to calculate the effective diffusivity from literature data on the drying of tobacco lamina and sliced celery. The applicability of the method was confirmed by good agreement of calculated and experimental data. It was found that the effective diffusivity for tobacco lamina rises sharply from practically zero at the beginning of drying to a maximum of 9·10^{− 10} m²/s at 89 s, and then gradually decays with time of drying. The same trend was found for sliced celery, but the maximum of the effective diffusivity (1.6·10^{− 7} m²/min) was attained at 56 s. The exact definition of the effective diffusivity vs. drying time identifies two phases of drying: the first phase was characterized by the rising intensity of drying with the maximum at the end of this phase, followed by the phase of slow decrease in the intensity of moisture removal. The rising intensity of drying observed during the initial phase of the process can be explained by warming up of the dried material during the initial phase of drying.