Successive pattern classification based on test feature classifier and its application to defect image classification

A novel successive learning algorithm based on a Test Feature Classifier is proposed for efficient handling of sequentially provided training data. The fundamental characteristics of the successive learning are considered. In the learning, after recognition of a set of unknown data by a classifier, they are fed into the classifier in order to obtain a modified performance. An efficient algorithm is proposed for the incremental definition of prime tests which are irreducible combinations of features and capable of classifying training patterns into correct classes. Four strategies for addition of training patterns are investigated with respect to their precision and performance using real pattern data. A real-world problem of classification of defects on wafer images has been dealt with by the proposed classifier, obtaining excellent performance even through efficient addition strategies.     

A study of quantum interference fluctuations in deep sub-μmMOSFET's under cryogenic conditions

We investigated the phase coherence length, l_φ, in large Si-MOSFET's fabricated using current process technology, with a particular emphasis on highly doped silicon substrates, and then studied the effects of quantum conductance fluctuations in deep sub-μm MOSFET's, with channel length comparable to l_φ. We identified, in a 0.2 μm MOSFET, universal conductance fluctuations in the strong inversion regime and conductance fluctuations due to variable range hopping in the weak inversion regime. The drain bias dependence of these fluctuations indicates clearly that they become a serious concern only at drain voltages lower than 10 mV. Therefore, even if the wave nature of electrons results in quantum conductance fluctuations, it will not lead to a limitation on device miniaturization in future Si-ULSI's     

β-phase transformation and thermoelectric power in FeSi2 and Fe2Si5 based alloys containing small amounts of Cu

The effects of Cu addition on the β phase formation rate and the thermoelectric power in various FeSi₂ and Fe₂Si₅ based alloys was examined. The peritectoid reaction (a+→β) in FeSi₂ alloys was initially enhanced by the addition of Cu but it became slower for longer annealing times. The retained metallic ε was harmful for the thermoelectric power. The inherent thermoelectric properties of (FeSi₂)_99−XMn₁Cu_X (X=0–1.O at.%), (FeSi₂)_99−X Co₁Cu_X (X=0–1.0 at.%) alloys were attained after the elimination of ε. In the case of eutectoid reaction (→β+Si). Differential thermal analysis, X-ray diffraction and microscopic observation clearly confirmed that the eutectoid reaction rate was drastically enhanced by the addition of a small amount of Cu and its rate decreased with decreasing Cu content. The rate also depends on the annealing temperature and reached a maximum at about 1073 K for most alloys. The addition of only 0.1 at.% Cu was still very effective even in Mn or Co doped alloys. The thermoelectric power of these alloys increased very quickly with annealing time. Their final values decreased with Cu content and saturated at 0.2 at.% Cu. The value of the 0.1 at.% Cu added alloy was higher than that of both the conventional p- and a-type FeSi₂ based alloys. These results suggest that the Fe₂Si₅ alloys with a small amount of Cu may be attractive as new thermoelectric materials.     

Total emission control for managing regional water quality: a game theoretic decision process

A simple game theoretic model with hierarchical structure is developed for regulating the total amount of pollutant discharged by all the effluent producers in a given region, where the allowable limit on total emission, that is the environmental carrying capacity, is assumed to be given. It is proposed that the total emission is to be allocated amongst all the polluters so as to maximize the social benefit, and that the social benefit is to be allocated through taxation on the companies under the criterion of impartiality. The concept of impartiality is defined as equal individual satisfaction for the polluters who cooperate. This concept is applied to solve the characteristic function game. A numerical example is included.     

Detection of light-absorbing iron oxide particles using a modified single-particle soot photometer

Copyright © 2016 American Association for Aerosol Research     

Isotope dilution analysis of Se in human blood serum by using high-power nitrogen microwave-induced plasma mass spectrometry coupled with a hydride generation technique

To establish a method for sensitive, accurate, and precise determination of Se in real samples, isotope dilution analysis using high-power nitrogen microwave-induced plasma mass spectrometry (N2 MIP-IDMS) was conducted. In this study, freeze-dried human blood serum (Standard Reference Material, NIES No. 4) provided by NIES (National Institute for Environmental Studies) was used as a real sample. The measured isotopes of Se were 78Se and 80Se which are the major isotopes of Se. The appropriate amount of a Se spike solution was theoretically calculated by using an error multiplication factor (F) and was confirmed experimentally for the isotope dilution analysis. The mass discrimination effect was corrected for by using a standard Se solution for the measurement of Se isotope ratios in the spiked sample. However, the sensitivity for the detection of Se was not so good and the precision of the determination was not improved (2-3%) by N2 MIP-IDMS with use of the conventional nebulizer. Therefore, a hydride generation system was connected to N2 MIP-IDMS as a sample introduction system (HG-N2 MIP-IDMS) in order to establish a more sensitive detection and a more precise determination of Se. A detection limit (3 sigma) of 10 pg mL-1 could be achieved, and the RSD was less than 1% at the concentration level of 5.0-10.0 ng mL-1 by HG-N2 MIP-IDMS. The analytical results were found to be in a good agreement with those obtained by the standard addition method using conventional Ar ICPMS.     

Transient power characteristics of balanced three‐phase variable inductance

Inductance has a particularly important role in power circuits. The authors have proposed the Variable Active–Passive Reactance (VAPAR), which can produce a virtual variable inductance. In this paper, the transient power characteristics of a balanced three‐phase variable inductance are analyzed using the instantaneous active and reactive power theorem. With an ideal voltage source or an ideal current source connected to a balanced three‐phase variable inductance, the transient power characteristics are analyzed theoretically, and then the results are verified through simulations. In the other cases, it is difficult to solve for the transient power, but it is shown that the transient power characteristics can easily be estimated from the results of simulations in which an ideal voltage source or an ideal current source is connected to a balanced three‐phase variable inductance. Experiments are also carried out to verify the theoretical analysis and simulations. © 2001 Scripta Technica, Electr Eng Jpn, 136(2): 49–57, 2001     

Solid-State Polymerization Behavior of 1,3-Bis(3-quinolyl)-1,4-butadiyne

Summary 1,4-Bis(3-quinolyl)-1,3-butadiyne (DQ) is known to be polymerized in solid state to give the corresponding polydiacetylene. However, the polymer yield of DQ bulk crystals is low. Thus, we prepared several types of DQ crystals by different procedures to find the reason for the low polymer yield. We found three modifications of DQ crystals and they were evaluated by spectroscopic measurements and X-ray diffraction. DQ bulk crystals (Crystal I) and thermally grown DQ crystals on Crystal I (Crystal II) have the same structure classified to Type A, which gives regular polydiacetylene structure in low polymer yields. DQ crystals grown on glass plates by sublimation (Crystal III) belong to the second modification of Type B. DQ nanocrystals prepared by the reprecipitation method (Crystal IV) are Type B, and thermally grown DQ crystals on nanocrystals (Crystal V) are Type C. Crystals of Types B and C could be polymerized in low yields without showing excitonic absorption of polydiacetylene indicating irregular polymerization other than 1,4-addition. For crystals of Type A, we found that the 1,4-addition polymerization proceeded only in near-surface portions of the crystals. It can be plausibly explained that mobile monomers in the near-surface portion are only able to take part in their polymerization, resulting in low polymer conversion.