Model search is a crucial component of data analytics pipelines, and this laborious process of choosing an appropriate learning algorithm and tuning its parameters remains a major obstacle in the widespread adoption of machine learning techniques. Recent effortsaiming to automate this process have assumed model training itself to be a black-box, thus limiting the effectiveness of such approaches on large-scale problems. In this work, we build upon these recent efforts. By inspecting the inner workings of model training and framing model search as bandit-like resource allocation problem, we present an integrated distributed system for model search that targets large-scale learning applications. We study the impact of our approach on a variety of datasets and demonstrate that our system, named GHOSTFACE, solves the model search problem with comparable accuracy as basic strategies but an order of magnitude faster. We further demonstrate that GHOSTFACE can scale to models trained on terabytes of data across hundreds of machines.

关键词：型号搜索;数据分析;分布式系统模型搜索;速度模型搜索

查看摘要 索取原文[13页]

Various RF circuit blocks implemented by using commercially available MEMS digital capacitor arrays are presented for reconfigurable RF systems. The designed circuit blocks are impedance-matching network, tunable bandpass filter, and VSWR sensor. The frequency range of the designed circuits is 0.4-4GHz. The MEMS digital capacitor arrays that are employed in the designs have built-in dc-to-dc voltage converter and serial interface significantly simplifying the control circuitry. The RF circuit blocks are suitable to low-cost, high-level of integration, thanks to the commercially available parts and standard RF packaging technologies.

关键词：电容器；微机电系统；带通滤波器；现场可编程门阵列

查看摘要 索取原文[5页]

Cancer informatics is a multidisciplinary field of research. It includes oncology, pathology, radiology, computational biology, physical chemistry, computer science, information systems, biostatistics, machine learning, artificial intelligence (AI), data mining and many others. Machine learning (ML) offers potentially powerful tools, intelligent methods, and algorithms that can help in solving many medical and biological problems. The variety of ML algorithms enable the design of a robust techniques and new methodologies for managing, representing, accumulating, changing, discovering ,and updating knowledge in cancer-based systems. Moreover it supports learning and understanding the mechanisms that will help oncologists, radiologists and pathologists to induce knowledge from cancer information databases. This paper presents the research results of the author and his colleagues that have been carried out in recent years on using machine learning in cancer informatics. In addition the talk discusses several directions for future research.

关键词：机器学习；癌症信息学；基于案例的推理（CBR）；本体工程

查看摘要 索取原文[14页]

Rodent baits containing systemic insecticides are potential tools to interrupt the cycle of transmission of Leishmania parasites by killing sand flies that take bloodmeals from rodents as adults. Bio-indicators that can be used in conjunction with rodent-targeted sand fly control methods also have been developed to demonstrate that the insecticide treatments are reaching the targeted life stages of sand flies and to quantify the effect of the insecticide treatments on sand fly populations. This article presents new results from a field study on the use of rodent bait containing a systemic insecticide in Kenya. The objective of this field study was to incorporate the fluorescent dye rhodamine B into rodent baits to determine the level of blood feeding by the sand fly Phlebotomus duboscqi on targeted rodents, and to demonstrate the effect of rodent bait containing the systemic insecticide ivermectin on blood-feeding adult females of P. duboscqi. Over 50of the blood- fed females of P. duboscqi collected at sites that were treated with rodent baits containing rhodamine B alone were positive for the presence of rhodamine B while no blood-fed females of P. duboscqi collected at the sites treated with rodent baits containing rhodamine B plus ivermectin were positive for the presence of rhodamine B. The results of this field trial constitute proof of concept for the targeted control of an epidemiologically significant portion of the population of the sand fly vector of Leishmania major, and demonstrate the potential for the interruption of the transmission of L. major using applications of systemic insecticide-treated rodent baits.

关键词：医药；双翅目；杀虫剂；矢量控制；伊维菌素

查看摘要 索取原文[7页]

This work outlines a new approach for online learning from imprecise data, namely, fuzzy set based evolving modeling (FBeM) approach. FBeM is an adaptive modeling framework that uses fuzzy granular objects to enclose uncertainty in the data. The FBeM algorithm is data flow driven and supports learning on an instance-per-instance recursive basis by developing and refining fuzzy models on-demand. Structurally, FBeM models combine Mamdani and functional fuzzy systems to output granular and singular approximations of nonstationary functions. In general, approximand functions can be time series, decision boundaries between classes, and control and regression functions. Linguistic description of the behavior of the system over time is provided by information granules and associated rules. An application example on a reactive control problem, underlining the complementarity of Mamdani and functional parts of the model, illustrates the usefulness of the approach. More specifically, the problem concerns sensor-based robot navigation and localization. In addition to precise singular output values, granular output values provide effective robust obstacle avoidance navigation.

关键词：数据流；自适应建模框架；机器人导航；回归函数

查看摘要 索取原文[15页]

关键词：超光速；集群计算框架；分布式文件系统

查看摘要 索取原文[17页]

Energy-saving is one of the inevitable problems of the routing design in WSN,while Data Fusion technology is widely utilized in energy constraint WSN to reduce the amount of messages exchanged between sensor nodes.This paper proposes a new algorithm based on Integrated Genetic and BP Neural Network(IGBP),IGBP uses the global search capability of GA to remedy the deficiency of BP artificial neural network.First,IGBP generates the best individuals in different networks by GA algorithm.Then it chooses the most optimize individual measure by Mean Squared Error to construct the BP network which was supplied to train of the WSN.Using the optimize individual nodes as initialization value training the BP network,it will enhance the learning rates of convergence and avoid falling into the local minimums.The simulation results show that the IGBP algorithm has made great progress in balancing the consumption of energy so as to prolong the network lifetime.

关键词：数据融合；遗传算法；BP人工神经网络

查看摘要 索取原文[5页]

This study focuses on the feasibility of using solar energy on Department of Defense (DoD) installations located throughout the Middle East. These DoD installations are currently using electricity generated from the local power grid at commercial rates, or generated by contractor-provided diesel generators. As an alternative, the effective use of solar energy may result in considerable savings and operational efficiency. In contrast to oil prices, the cost of renewable energy has been declining rapidly in recent years. The cost of solar panels, for example, has decreased by more than 60 percent since 2009 (Reichart, 2011). The aim of this paper is to carry out an analysis of the use of solar photovoltaic (PV) panels at DoD installations with respect to their cost-effectiveness in installation, operation, maintenance, security, sustainability, and reliability. We will examine meteorological data that provide the average monthly sunlight levels for different geographical areas, information on the effectiveness of the commercial solar PV panels currently available, electricity consumption levels for DoD installations at different times of the day, and current cost structures. Noneconomic attributes of solar installations also need to be considered to properly assess the feasibility of solar power as a substitute for current sources of electricity. The research is limited to evaluating the feasibility of using solar PV panels for electricity generation in forward operating bases. We have analyzed a trend in levelized cost of electricity (LCOE) for solar PV panels and have come to the conclusion that solar PV panels will become competitive with grid energy around 2020, provided that current growth projections in installed capacity and a corresponding decrease in the cost of solar PV-generated electricity continue. In light of our findings, we are confident that solar PV has the potential to compete with grid electricity in the near future.

关键词：电力生产；光电细胞（半导体）；太阳能电池板；太阳能

查看摘要 索取原文[57页]

Oxygen plays a fundamental role in nature and industrial process,the fast on-line oxygen concentration detection is of great significance.Tunable diode laser absorption spectroscopy (TDLAS) is capable high selective,high sensitive,non-intrusive,this technique exploits the diode laser tunable and narrow bandwidth characteristics,by selecting one absorption line of the target gas to avoid interference from other gas and realize oxygen concentration fast on-line detection.In this paper,we use vertical cavity surface-emitting laser (VCSEL) as the light source,adopting the wavelength modulation spectroscopy and the harmonic detection technology to measure the second harmonic signal and realize oxygen concentration fast on-line detection.The detection concentration range between 0.1％ to 40％ and detection accuracy is 0.1％.

关键词：氧传感器；TDLAS；WMS VCSEL

查看摘要 索取原文[5页]

This study measured inflammatory mediators in tracheal secretions and plasma of subjects with severe multiple blunt trauma for 4 days post injury. The goal was to determine if biomarkers in tracheal fluid could be used to discriminate between subjects who developed acute lung injury (ALI) after Day 1 from those who did not develop ALI. The tracheal samples were analyzed by enzyme-linked immunosorbent assay for quantities of tumor necrosis factor (TNF), interleukin 1 beta (IL-1 ), IL-6, IL-8, IL-10, C-reactive protein, pentraxin 3 (PTX3), and activated complement 5 daily for 4 days. The diagnosis of ALI or acute respiratory distress syndrome was made using the international consensus criteria. Comparisons were made from samples collected on the first 24 hours following injury (Day1) between the No ALI group (n=15) and the ALI group (n=6). The levels of all eight mediators on Day 1 were higher in the No ALI group vs. levels in the group that developed ALI after Day 1. The levels of TNF, IL-1 , and PTX3 were statistically higher in the No ALI group. The significant differences in biomarkers in tracheal fluid demonstrate that these markers are potentially predictive of which trauma subjects are at risk for ALI.

关键词：肺；创伤和损伤;化验；血浆；细胞因子

查看摘要 索取原文[25页]