Bias and Controversy in Evaluation Systems

Evaluation is prevalent in real-life. With the advent of Web 2.0, online evaluation has become an important feature in many applications that involve information (e.g., video, photo, audio) sharing, and social networking (e.g., blogging). In these evaluation settings, a set of reviewers assign scores to a set of objects. As part of evaluation analysis, we want to obtain fair reviews for all the given objects. However, the reality is that reviewers may deviate in their scores assigned to the same object, due to the potential "bias" of reviewers or "controversy" objects. The statistical approach of averaging deviations to determine bias and controversy assumes that all reviewers and objects should be given equal weight. In this paper, we look beyond this assumption and propose an approach based on the following observations: (1) evaluation is "subjective", as reviewers and objects have varying bias and controversy respectively, and (2) bias and controversy are mutually dependent. These observations underlie our proposed reinforcement-based model to determine bias and controversy simultaneously. Our approach also quantifies "evidence", which reveals the degree of confidence with which bias and controversy has been derived. This model is shown to be effective by experiments on real-life and synthetic datasets.     

Variational learning from implicit bandit feedback

Machine Learning - Recommendations are prevalent in Web applications (e.g., search ranking, item recommendation, advertisement placement). Learning from bandit feedback is challenging due to the...     

Surface topography and tribological properties of coatings prepared from microparticle size polyurethane dispersions studied by atomic force microscopy

The surface topography and mechanical properties of coatings prepared using large particle size polyurethane dispersions (PUD) are investigated using atomic force microscopy (AFM) imaging, AFM-based force measurements, and friction force microscopy. PUD coatings, which are prepared from dispersions containing particles of micron size, have surface roughness of 250–300 nm and waviness of 2.5–3 μm resulting from the particle size. The surface moduli of the PUD coatings are varied by tuning the ratio of hard-to-soft segmentation in the polyurethanes and are found to be between 40 and 100 MPa. The friction coefficient obtained in the study is found to be correlated with both the surface modulus of the coatings and the adhesion between the probe and the samples and is well in line with the perceived feel of an experienced human panel. The data are very well behaved and clearly show the utility of this technique in characterizing these types of surfaces.     

Synthesis and characterization of cytidine derivatives that inhibit the kinase IspE of the non-mevalonate pathway for isoprenoid biosynthesis

The enzymes of the non-mevalonate pathway for isoprenoid biosynthesis are attractive targets for the development of novel drugs against malaria and tuberculosis. This pathway is used exclusively by the corresponding pathogens, but not by humans. A series of water-soluble, cytidine-based inhibitors that were originally designed for the fourth enzyme in the pathway, IspD, were shown to inhibit the subsequent enzyme, the kinase IspE (from Escherichia coli). The binding mode of the inhibitors was verified by co-crystal structure analysis, using Aquifex aeolicus IspE. The crystal structures represent the first reported example of a co-crystal structure of IspE with a synthetic ligand and confirmed that ligand binding affinity originates mainly from the interactions of the nucleobase moiety in the cytidine binding pocket of the enzyme. In contrast, the appended benzimidazole moieties of the ligands adopt various orientations in the active site and establish only poor intermolecular contacts with the protein. Defined binding sites for sulfate ions and glycerol molecules, components in the crystallization buffer, near the well-conserved ATP-binding Gly-rich loop of IspE were observed. The crystal structures of A. aeolicus IspE nicely complement the one from E. coli IspE for use in structure-based design, namely by providing invaluable structural information for the design of inhibitors targeting IspE from Mycobacterium tuberculosis and Plasmodium falciparum. Similar to the enzymes from these pathogens, A. aeolicus IspE directs the OH group of a tyrosine residue into a pocket in the active site. In the E. coli enzyme, on the other hand, this pocket is lined by phenylalanine and has a more pronounced hydrophobic character.     

Series-resonant single-phase AC-DC power supply with control ofreactive power

It is shown that the resolution of the high-frequency input current of the series-resonant converter leads to the possibility of accurately controlling the waveform of the source current. In combination with the accurate positioning of the high-frequency current pulses, it is possible to influence the exchange of power with the source by introducing active filtering. A power factor equal to one would lead to optimal operation. However, the zero crossing of the AC source voltage, introduces the need to store energy. Bulky low-frequency storage elements increase the specific volume and weight of the converter, while the exchange of energy with storage elements decreases the overall efficiency. The energy stored in the resonant circuit is fundamentally too low to solve this problem. The stored energy in the output capacitor necessary to decrease the high-frequency ripple voltage is available for these purposes. To meet the optimal conditions for the power factor at the source, a bipolar flow of energy is required     

Shared-storage auction ensures data availability

Most current e-auction systems are based on the client-server architecture. Such centralized systems provide a single point of failure and control. In contrast, peer-to-peer systems permit distributed control and minimize individual node and link failures' impact on the system. The shared-storage-based auction model described decentralizes services among peers to share the required processing load and aggregates peers' resources for common use. The model is based on the principles of local computation at each peer, direct inter-peer communication, and a shared storage space.     

Endotoxin release and cytokine production in acute and chronic meningococcaemia

Muscle activity produces an electrical signal termed the myo-electric signal (MES). The MES is a useful clinical tool, used in diagnostics and rehabilitation. This signal is typically stored in 2 bytes as 12-bit data, sampled at 3 kHz, resulting in a 6 kbyte s-1 storage requirement. Processing MES data requires large bit manipulations and heavy memory storage requirements. Adaptive differential pulse code modulation (ADPCM) is a popular and successful compression technique for speech. Its application to MES would reduce 12-bit data to a 4-bit representation, providing a 3:1 compression. As, in most practical applications, memory is organised in bytes, the realisable compression is 4:1, as pairs of data can be stored in a single byte. The performance of the ADPCM compression technique, using a real-time system at 1 kHz, 2 kHz and 4 kHz sampling rates, is evaluated. The data used include MES from both isometric and dynamic contractions. The percent residual difference (PRD) between an unprocessed and processed MES is used as a performance measure. Errors in computed parameters, such as median frequency and variance, which are used in clinical diagnostics, and waveform features employed in prosthetic control are also used to evaluate the system. The results of the study demonstrate that the ADPCM compression technique is an excellent solution for relieving the data storage requirements of MES both in isometric and dynamic situations.     

Source-sided power flow control of a single-phase AC-to-DC series-resonant power supply

Contents The objective of an AC-to-DC converter is to control the output waveform. The energy emanated from the source is adapted by passive filters, which is often a compromise between volume and effective use. A classical rectifierfilter network presents a major disadvantage: a power factor less than one which is caused by distortion of the source current.The resolution of the high-frequency input current of the series-resonant converter introduces the possibility of controlling the waveform of the source current accurately. In combination with the accurate positioning of the high-frequency current pulses, it is possible to influence the exchange of power with the source by the introduced method of active filtering. A power factor equal to one would lead to an optimal operation. The method introduced can select any value for the power factor, even for distorted voltage waveforms.However the zero-crossing of the AC source voltage introduces the necessity of storing energy. Bulky low-frequency storage elements increase the specific volume and weight of the converter while the exchange of energy with storage elements decreases the overall efficiency. The energy stored in the resonant circuit is fundamentally too low to solve this problem. The stored energy in the output capacitor, necessary to decrease the high-frequency ripple voltage, is available for these purposes. To meet the optimal conditions for the power factor at the source, a bipolar flow of energy is required.The theory of single-phase AC-to-DC series-resonant power conversion with controllable reactive power is explained. The results of simulation and experiment will establish the extensive possibilities of the proposed systems.

---

Netzseitige Leistungsflußregelung eines einphasigen AC-DC serienresonanten Leistungsumformers

Übersicht Die Aufgabe eines AC-DC Umrichters ist es, die Spannungsform des Ausganges regelungstechnisch festzulegen. Die von der Speisequelle entnommene Energie wird über passive Filter ausgekoppelt, die stets einen Kompromiß zwischen Raumbedarf und erzielter Wirkung bilden. Ein klassischer Gleichrichter liefert, besonders, wenn er gesteuert ist, stets einen netzseitigen Leistungsfaktor, der kleiner eins ist. Dies ist durch den verformten und in der Phase verschobenen Netzstrom bedingt.Die hohe Auflösung des hochfrequenten Eingangsstromes eines serienresonanten Umrichters bietet die Möglichkeit, die netzseitige Stromform sehr genau zu regeln. In Verbindung mit der genauen Positionierung der hochfrequenten Strompulse ist es möglich, die momentan zugeführte Energie mit der vorgestellten Methode der aktiven Filterung zu beeinflussen. Ein geregelter Leistungsfaktor von eins führt so zu einem optimalen Betriebszustand. Die vorgesehene Methode gestattet es, jeglichen Leistungsfaktor einzustellen, selbst bei einer pulsierenden Spannungszeitfunktion.Der Nulldurchgang der Wechselspannung des Netzes zeigt die Notwendigkeit, Energie zwischenzeitlich zu speichern. Speicher für niedrige Frequenzen erhöhen das Gewicht und das Volumen des Umrichters, gleichzeitig läßt der Austausch von Energie mit Energiespeichern den Wirkungsgrad sinken.Die gespeicherte Energie in einem hochfrequenten Resonanzkreis ist also grundsätzlich zu gering, um dieses Problem zu lösen. Die gespeicherte Energie dagegen in den erforderlichen Ausgangsfilter-Kondensatoren, die zur Glättung der Spannung dienen, ist für diesen Zweck gut geeignet. Um optimale Bedingungen hinsichtlich des netzseitigen Leistungsfaktors zu erhalten, ist ein Leistungsfluß in beiden Richtungen erforderlich.Die Theorie eines einphasigen AC-DC serienresonanten Umrichters mit regelbarem Leitungsfaktor wird erläutert. Simulationen und Meßresultate einer derartigen Ausführung werden vorgestellt, die die weiten Möglichkeiten einer Leistungsflußregelung verdeutlichen.

     

Variable-speed generation with the series-resonant converter

Results of study and measurements are presented for a variable-speed generation (VSG) system that uses a series-resonant converter to excite the rotor windings of a doubly-fed generator. The capabilities of this system include maximum-efficiency operation and limited converter rating for a speed range from stand-still (zero RPM) to well beyond synchronous speed, as well as flexibility for line-voltage control, reactive power generation, and synchronization at zero RPM. Measurements of the harmonic distortion are provided showing the superiority of this converter as compared with a VSG system using a conventional cycloconverter or a DC-link converter. A convenient tool for analyzing the operating modes of the system is discussed that can be developed under varying resource conditions