Mechanism of production of intestinal secretion by negative luminal pressure

The relationships between luminal hydrostatic pressure and fluid transport by dog jejunum in vivo studied as a sheet in the Wells clamp were compared with quantitative predictions from a model proposed for the mechanism of the secretion produced by elevated venous pressure. According to the model, the secretion produced by increasing venous pressure and the secretion produced by negative luminal pressure are both passive filtrates contingent on a transepithelial pressure of a few centimeters of H2O. We consider that the agreement between the observed and predicted responses to luminal pressure provides strong support for the model. In particular, a) the observations displayed a predicted gross asymmetry in rates of fluid transfer with isotonic fluids depending on whether the luminal pressure was positive or negative; b) the observed magnitude of the negative luminal pressure required for the onset of secretion agreed with predictions; and c) the secretion contained significant amounts of protein at about 25% of the plasma concentration.     

Phase retrieval from subdivision of the focal plane with a lenslet array

A phase retrieval algorithm derived from subdivision of the complex field at the focal plane is proposed. This subdivision is achieved with a lenslet array at the focal plane in a manner similar to the pyramid wave-front sensor. The phase retrieval algorithm significantly improves the wave-front estimate that can be attained as a linear combination of the aperture images. This phase retrieval algorithm also avoids the twin-image stagnation problem inherent in phase retrieval and phase retrieval in conjunction with the Shack-Hartmann sensor.     

An Empirical Study of Two Approaches to Sequence Learning for Anomaly Detection

This paper introduces the computer security domain of anomaly detection and formulates it as a machine learning task on temporal sequence data. In this domain, the goal is to develop a model or profile of the normal working state of a system user and to detect anomalous conditions as long-term deviations from the expected behavior patterns. We introduce two approaches to this problem: one employing instance-based learning (IBL) and the other using hidden Markov models (HMMs). Though not suitable for a comprehensive security solution, both approaches achieve anomaly identification performance sufficient for a low-level focus of attention detector in a multitier security system. Further, we evaluate model scaling techniques for the two approaches: two clustering techniques for the IBL approach and variation of the number of hidden states for the HMM approach. We find that over both model classes and a wide range of model scales, there is no significant difference in performance at recognizing the profiled user. We take this invariance as evidence that, in this security domain, limited memory models (e.g., fixed-length instances or low-order Markov models) can learn only part of the user identity information in which we're interested and that substantially different models will be necessary if dramatic improvements in user-based anomaly detection are to be achieved.     

Comparison of two JAERI phantoms and the problems discovered

During the course of an intercomparison exercise it was possible to compare two JAERI phantoms with each other by using a multi-energy photon emitting lung set (241Am/152Eu). One belonged to the IAEA (Vienna), the other belonged to the Human Monitoring Laboratory (Ottawa). The intercomparison of the phantoms showed that they were statistically distinct from each other, although the differences were small. The counting efficiencies varied from each other by about 4% at 17 keV and 2% at photon energies above 17 keV. It was concluded that these difference were either due to small variations of chest wall thickness during the manufacturing process or positioning errors. The intercomparison also revealed a serious problem with one of the overlay plates of the HML's phantom. The adipose mass fraction of the overlay plate was found to be much greater (approximately 40%) than the manufacturer's stated value (10%).     

Using the Curriculum Vita to Study the Career Paths of Scientists and Engineers: An Exploratory Assessment

In this paper we assess the utility of the curriculum vita (CV) as a data source for examining the career paths of scientists and engineers. CVs were obtained in response to an email message sent to researchers working in the areas of biotechnology and microelectronics. In addition, a number of CVs were obtained "passively" from a search of the Internet. We discuss the methodological issues and problems of this data collection strategy and the results from an exploratory analysis using OLS regression and event history analysis. In sum, despite difficulties with coding and variation in CV formats, this collection strategy seems to us to hold much promise.     

Detecting faults in heterogeneous and dynamic systems using DSP and an agent-based architecture

This paper demonstrates the use of multi-agent systems (MAS), firstly as a modelling technique for dynamic physical systems and secondly as the basis for a generic and powerful diagnostic system, which can support heterogeneous distributed systems. First an overview of the diagnostic techniques including those offered by the two communities fault detection and isolation (FDI ) and DX (based on intelligent techniques) is given. The use of digital signal processing (DSP) as a significant technique for improved fault diagnosis is illustrated. A rule-based engine is used to control the behaviours of the agents and also as a tool for diagnosis. Finally, the integration of DSP agents and the rule-based engine into MAS is demonstrated using a real-life application, a class-AB amplifier (a power electronic circuit). It is shown that the integration of DSP agents and rules into MAS provides a powerful tool for prognosis and for detection of abrupt (short and open circuit) and incipient faults.     

Bilingual LSA-based adaptation for statistical machine translation

We propose a novel approach to cross-lingual language model and translation lexicon adaptation for statistical machine translation (SMT) based on bilingual latent semantic analysis. Bilingual LSA enables latent topic distributions to be efficiently transferred across languages by enforcing a one-to-one topic correspondence during training. Using the proposed bilingual LSA framework, model adaptation can be performed by, first, inferring the topic posterior distribution of the source text and then applying the inferred distribution to an n-gram language model of the target language and translation lexicon via marginal adaptation. The background phrase table is enhanced with the additional phrase scores computed using the adapted translation lexicon. The proposed framework also features rapid bootstrapping of LSA models for new languages based on a source LSA model of another language. Our approach is evaluated on the Chinese–English MT06 test set using the medium-scale SMT system and the GALE SMT system measured in BLEU and NIST scores. Improvement in both scores is observed on both systems when the adapted language model and the adapted translation lexicon are applied individually. When the adapted language model and the adapted translation lexicon are applied simultaneously, the gain is additive. At the 95% confidence interval of the unadapted baseline system, the gain in both scores is statistically significant using the medium-scale SMT system, while the gain in the NIST score is statistically significant using the GALE SMT system.     

嵌入式USB主机

1 USB主机控制器 1.1 对于主板 早期的USB主控制器通过PCI总线与主板连接,现今的USB功能则内建在主板芯片组中.这些基于PCI总线的控制器可处理大量USB事务,例如将PC和多个USB设备之间传输的数据包排序和分级.     

The External Sourcing of Technological Knowledge by US Pharmaceutical Companies: Strategic Goals and Inter-organizational Relationships

This study of the organization of the discovery function by large US pharmaceutical companies (LPCs) examines an important knowledge acquisition strategy—external sourcing of compounds and technologies. Through a critical examination of the “capabilities” thesis in strategic management and of the theoretical conceptualization in organization studies of innovation networks, we undertake an in-depth qualitative analysis of the network relationships LPCs establish with biotechnology firms/public research laboratories. We additionally examine the motivations and degree of strategic intent of R&D managers involved in external knowledge sourcing. The paper identifies the tensions and contradictions in network relationships and indicates how these lead to changes in knowledge sourcing. This qualitative analysis is placed in its industry and technology context. This reveals both the pressures towards and the trends in external knowledge sourcing, as compared with in-house discovery. Extensive interviews with US LPCs and biotechnology firms provide a rare glimpse of how some of the most important actors in global innovation networks handle a significant new innovation strategy.