PAC-learning in the presence of one-sided classification noise

We derive an upper and a lower bound on the sample size needed for PAC-learning a concept class in the presence of one-sided classification noise. The upper bound is achieved by the strategy “Minimum One-sided Disagreement”. It matches the lower bound (which holds for any learning strategy) up to a logarithmic factor. Although “Minimum One-sided Disagreement” often leads to NP-hard combinatorial problems, we show that it can be implemented quite efficiently for some simple concept classes like, for example, unions of intervals, axis-parallel rectangles, and TREE(2,n,2,k) which is a broad subclass of 2-level decision trees. For the first class, there is an easy algorithm with time bound O(m logm). For the second-one (resp. the third-one), we design an algorithm that applies the well-known UNION-FIND data structure and has an almost quadratic time bound (resp. time bound O(n ² m logm)).     

Identifizierung von molekularen Phospholipid-Species aus Schweineherz durch Kombination chromatographischer Techniken

Identification of Molecular Species of Phospholipids from Pig Heart by Combining Chromatographic Techniques. Glycero-phospholipids from subcellular organelles and membranes from pig heart were divided into subclasses. The molecular species within the main subclasses were analyzed. The phosphorus moiety had been removed by phospholipase-C and the acetylated molecules were separated in diacyl-, alkylacyl- and alkenylacyl-glycerols by thin-layer chromatography (TLC). For further identification the subclasses from the first fractionation by TLC were separated by silver ion chromatography according to their polarity. A small portion of the acetylated diradyl-species was transesterified so that the fatty acid composition could be assessed as methyl esters. The acetylated molecular species derived from glycerophospholipids were chromatographed on WCOT fused silica columns with TAP as stationary phase according to molecular weight and unsaturation. Gaschromatography (GC) of fatty acid methyl esters and dimethylacetals from plasmalogens was executed on DEGS-columns or on WCOT fused silica columns with CP-Sil 88 as stationary phase. The distribution of molecular species within a phospholipid class (PC or PE) was rather similar for the diacyl-and the alkenylacyl molecules. But the latter are more volatile. In the case of phosphati-dylcholine the composition of C₁₆-C_18-:1; C₁₆-C_18:2 and C₁₈-C_18:2 was predominant for diacyl- as well as alkenylacyl-species. In the case of phosphatidylethanolamine the fatty acid composition of C₁₈-C_20:4 and C₁₈-C_18:2 were the most important species. The composition of fatty acids in diacyl-and alkenylacyl-species is of special interest as the metabolism of diacyl-glycerophospholipids is quite different from alkenylacyl-glycerophospholipids.     

Analysis of platinum adducts with DNA nucleotides and nucleosides by capillary electrophoresis coupled to ESI-MS: indications of guanosine 5'-monophosphate O6-N7 chelation

DNA is the ultimate target of platinum-based anticancer therapy. Since the N7 of guanine is known to be the major binding site of cisplatin and its analogues, adduct formation with model nucleotides, especially 2'-deoxyguanosine 5'-monophosphate (dGMP), has been studied in detail. During the last few years a coupled capillary eletrophoresis/electrospray-ionization mass spectrometry (CE/ESI-MS) method has been advantageously used in order to separate and identify platinum adducts with nucleotides in submillimolar concentrations in aqueous solutions. Beside the bisadduct, [Pt(NH(3))(2)(dNMP)(2)](2-) (NMP=2'-deoxynucleoside 5'-monophosphate), and the well-known monochloro and monohydroxo adducts, [Pt(NH(3))(2)Cl(dNMP)](-) and [Pt(NH(3))(2)(dNMP)OH](-), respectively, a third kind of monoadduct species with a composition of [Pt(NH(3))(2)(dNMP)](-) can be separated by CE and detected through the m/z values measured with ESI-MS. Different experimental setups indicate the existence of an O(6)-N7 chelate, whereas the formation of N7-alphaPO(4) macrochelates or dinuclear species is unlikely. Additionally, offline MS experiments with 2'-deoxyguanosine (dG) and stabilization of the controversially discussed O(6)-N7 chelate by oxidation with hydrogen peroxide support the assumption of the existence of O(6)-N7 chelation.     

Computer usage questionnaire: Structure, correlates, and gender differences

Computer usage, computer experience, computer familiarity, and computer anxiety are often discussed as constructs potentially compromising computer-based ability assessment. After presenting and discussing these constructs and associated measures we introduce a brief new questionnaire assessing computer usage. The self-report measure consists of 18 questions asking for the frequency of different computer activities and software usage. Participants were N = 976 high school students who completed the questionnaire and several covariates. Based on theoretical considerations and data driven adjustments a model with a general computer usage factor and three nested content factors (Office, Internet, and Games) is established for a subsample (n = 379) and cross-validated with the remaining sample (n = 597). Weak measurement invariance across gender groups could be established using multi-group confirmatory factor analysis. Differential relations between the questionnaire factors and self-report scales of computer usage, self-concept, and evaluation are reported separately for females and males. It is concluded that computer usage is distinct from other behavior oriented measurement approaches and that it shows a diverging, gender-specific pattern of relations with fluid and crystallized intelligence.     

Very fast and exact accumulation of products

The IFIP Working Group on Numerical Software and other scientists repeatedly requested that a future arithmetic standard should consider and specify an exact dot product (EDP) [The IFIP WG—IEEE 754R letter, dated September 4 (2007), The IFIP WG—IEEE P1788 letter, dated September 9 (2009)]. On 18 November 2009 the IEEE standards committee P1788 on interval arithmetic accepted a motion [Kulisch and Snyder (The exact dot product as basic tool for long interval arithmetic, passed on Nov 18, 2009 as official IEEE P1788 document)] for including the EDP into a future interval arithmetic standard. Actually the simplest and fastest way for computing a dot product is to compute it exactly. By pipelining, it can be computed in the time the processor needs to read the data, i.e., it comes with utmost speed. A hardware implementation of the EDP exceeds any approximate computation of the dot product in software by several orders of magnitude. By a sample illustration the paper informally specifies the implementation of the EDP on computers. While [Kulisch and Snyder (The exact dot product as basic tool for long interval arithmetic, passed on Nov 18, 2009 as official IEEE P1788 document)] defines what has to be provided, how to embed the EDP into the new standard IEEE 754, [IEEE Floating-Point Arithmetic Standard 754 (2008)] and how exceptions like NaN are to be dealt with, this article illustrates how the EDP can be implemented on computers. There is indeed no simpler way of accumulating a dot product. Any method that just computes an approximation also has to consider the relative values of the summands. This results in a more complicated method. The hardware needed for the EDP is comparable to that for a fast multiplier by an adder tree, accepted years ago and now standard technology in every modern processor. The EDP brings the same speedup for accumulations at comparable costs. In Numerical Analysis the dot product is ubiquitous. It is not merely a fundamental operation in all vector and matrix spaces. It is the EDP which makes residual correction effective. This has a direct and positive influence on all iterative solvers of systems of equations. The EDP is essential for fast long real and long interval arithmetic, as well as for assessing and managing uncertainty in computing. By operator overloading variable precision interval arithmetic is very easy to use. With it the result of every arithmetic expression can be guaranteed to a number of correct digits.     

Aluminum–germanium wafer bonding of (AlGaIn)N thin-film light-emitting diodes

Eutectic aluminum–germanium wafer bonding was used to fabricate (AlGaIn)N thin-film light-emitting diodes (LEDs). Wafer bonding was carried out on 2″ wafer level at a bond temperature of 470 °C using patterned Al bond pads on the GaN-on-sapphire LED epiwafer and plain Ge substrates. The microstructure of the joint formation was characterized via cross-section analysis using scanning electron microscopy and energy dispersive X-ray spectroscopy (EDX). Scanning acoustic microscopy was used to investigate the bond interface. The shear strength was determined to be 1–2 kN/cm². The formation of a liquid Al–Ge phase is evident from cross-section analysis and optical microscopy. During solidification, Al and Ge are separated into distinct phases again, which is revealed by EDX. The obtained bond is not free of micro-voids, yet it is mechanically stable and suited for the fabrication of thin-film LEDs by removing the sapphire substrate via laser lift-off, which is also demonstrated.     

New reactor dedicated to in operando studies of model catalysts by means of surface x-ray diffraction and grazing incidence small angle x-ray scattering

A new experimental setup has been developed to enable in situ studies of catalyst surfaces during chemical reactions by means of surface x-ray diffraction (SXRD) and grazing incidence small angle x-ray scattering. The x-ray reactor chamber was designed for both ultrahigh-vacuum (UHV) and reactive gas environments. A laser beam heating of the sample was implemented; the sample temperature reaches 1100 K in UHV and 600 K in the presence of reactive gases. The reactor equipment allows dynamical observations of the surface with various, perfectly mixed gases at controlled partial pressures. It can run in two modes: as a bath reactor in the pressure range of 1-1000 mbars and as a continuous flow cell for pressure lower than 10(-3) mbar. The reactor is connected to an UHV preparation chamber also equipped with low energy electron diffraction and Auger spectroscopy. This setup is thus perfectly well suited to extend in situ studies to more complex surfaces, such as epitaxial films or supported nanoparticles. It offers the possibility to follow the chemically induced changes of the morphology, the structure, the composition, and growth processes of the model catalyst surface during exposure to reactive gases. As an example the Pd(8)Ni(92)(110) surface structure was followed by SXRD under a few millibars of hydrogen and during butadiene hydrogenation while the reaction was monitored by quadrupole mass spectrometry. This experiment evidenced the great sensitivity of the diffracted intensity to the subtle interaction between the surface atoms and the gas molecules.     

High energy resolution five-crystal spectrometer for high quality fluorescence and absorption measurements on an x-ray absorption spectroscopy beamline

Fluorescence detection is classically achieved with a solid state detector (SSD) on x-ray absorption spectroscopy (XAS) beamlines. This kind of detection however presents some limitations related to the limited energy resolution and saturation. Crystal analyzer spectrometers (CAS) based on a Johann-type geometry have been developed to overcome these limitations. We have tested and installed such a system on the BM30B/CRG-FAME XAS beamline at the ESRF dedicated to the structural investigation of very dilute systems in environmental, material and biological sciences. The spectrometer has been designed to be a mobile device for easy integration in multi-purpose hard x-ray synchrotron beamlines or even with a laboratory x-ray source. The CAS allows to collect x-ray photons from a large solid angle with five spherically bent crystals. It will cover a large energy range allowing to probe fluorescence lines characteristic of all the elements from Ca (Z = 20) to U (Z = 92). It provides an energy resolution of 1-2 eV. XAS spectroscopy is the main application of this device even if other spectroscopic techniques (RIXS, XES, XRS, etc.) can be also achieved with it. The performances of the CAS are illustrated by two experiments that are difficult or impossible to perform with SSD and the complementarity of the CAS vs SSD detectors is discussed.     

Experimental and model assisted investigation of an operational strategy for the BPR under low influent concentrations

The behaviour of a pilot scale biological phosphorus removal process (BPR) of the alternating type was investigated during periods of low influent concentrations and increased hydraulic load. A process disturbance of this type result in an increase in the phosphate concentration level in the anoxic/aerobic reactors and in the plant effluent shortly after the influent wastewater returns to normal strength. The accumulation of phosphorus in the system was avoided by the addition of an external carbon source either to the influent or to the effluent from the anaerobic reactor in form of sodium acetate. With the help of such an addition, the internal carbon storage compounds could be maintained at a high level, which is shown by poly-hydroxy-alcanoates (PHA) measurements. Several levels of acetate addition were investigated experimentally in order to determine a minimal amount of internally stored carbon, which could ensure the stabilization of BPR during such dynamic influent conditions. Furthermore reduction of aeration time during periods of low influent concentrations was investigated. It was observed that BPR was stabilized by combining a reduction of aeration time with carbon source addition, which maintained the internal stored carbon at a higher level. This combined control action resulted in a desired high BPR activity when the normal strength of the influent wastewater was re-established. The failure of the BPR process was sometimes observed even when comparatively high concentrations of PHA could be detected and an identification of a minimal PHA level was not possible. During this investigation an extended version of the activated sludge model No. 2 (ASM2), which includes denitrification by phosphate accumulating organisms, is used for the detailed analysis of the experiments. The model predicted the phosphorus build-up after the process disturbance as well as the performance during the stabilized experiments. Assisted by the model, the investigations indicate that a PHA limitation is not the only factor affecting the recovery of the BPR process during periods of low influent concentrations.     

Process methodology for the small scale production of m6N5 purity zinc using a resistance heated vacuum distillation system

Ultra high purity Zn (m6N5) was obtained through triple vacuum distillation using an m4N5 Zn as input material. High-volatile impurities were removed from the zinc matrix by vacuum evaporation, while traces having lower volatility than the matrix remained in the residual material after finishing of the entire purification process. The time required for the removal of the main high-volatile impurity (Cd) was monitored using a connected gas source mass spectrometer. During distillation the Zn vapor of the distillate was condensed on the cap of a glassy carbon crucible which was water cooled by a movable copper block. The temperatures were optimized for the removal of the volatiles and for the distillation process using vapor pressure data. High resolution glow discharge mass spectrometry was applied for the analysis and purity evaluation of the distilled and input zinc. The analysis confirmed the reduction of the metallic impurities from 43 mg kg⁻¹ to 0.5 mg kg⁻¹ (m6N5) after three consecutive vacuum distillations. An increase in the grain size and a decrease in the micro-hardness were observed for the purified Zn material.