Evolving artificial neural network ensembles

Using a coordinated group of simple solvers to tackle a complex problem is not an entirely new idea. Its root could be traced back hundreds of years ago when ancient Chinese suggested a team approach to problem solving. For a long time, engineers have used the divide-and-conquer strategy to decompose a complex problem into simpler sub-problems and then solve them by a group of solvers. However, knowing the best way to divide a complex problem into simpler ones relies heavily on the available domain knowledge. It is often a manual process by an experienced engineer. There have been few automatic divide-and-conquer methods reported in the literature. Fortunately, evolutionary computation provides some of the interesting avenues to automatic divide-and-conquer methods. An in-depth study of such methods reveals that there is a deep underlying connection between evolutionary computation and ANN ensembles. Ideas in one area can be usefully transferred into another in producing effective algorithms. For example, using speciation to create and maintain diversity had inspired the development of negative correlation learning for ANN ensembles, and an in-depth study of diversity in ensembles. This paper will review some of the recent work in evolutionary approaches to designing ANN ensembles.     

基于DSP的电力数据采集平台的设计与研究

介绍了一种用于电力系统的数据采集系统.该系统采用数字信号处理器(DSP)和复杂可编程逻辑器件(CPLD)作为核心芯片,给出了主要部分的电路及其原理、功能,系统中使用CPLD实现基本逻辑,使系统有足够的冗余和灵活性,具有可靠性高、可升级等优点.     

Analyzing the performance and fairness of BitTorrent-like networks using a general fluid model

In this paper, a general fluid model is developed to study the performance and fairness of BitTorrent-like networks. The fluid model incorporates two important features previously isolated from system performance models, user settings with multiple groups and inter-group data exchange induced by the choking algorithm. Our numerical results point out some key parameters of performance, such as the staying time of seeders. Generally, selfish behavior does not receive equal performance degradation, and in some scenarios users have strong incentives of free-riding. We also find content delivery can be greatly deterred when malicious free-riders are overwhelming.     

New encapsulation development for fine pitch IC devices

The continuous reduction of chip size driven by the market demand has a significant impact on circuit design and assembly process of IC packages. Shrinking chip size and increasing I/O counts require finer bond pad pitch and bond pad size for circuitry layout. As a result, serious wire deflection during transfer molding process could make adjacent wires short, and this issue becomes more critical as a smaller wire diameter has to be applied for the finer pitch wire bonded IC devices.This paper presents a new encapsulation process development for 50 μm fine pitch plastic ball grid array package. Since reduced wire diameter decreases the bending strength of bonded wires significantly, wire deflection during molding process becomes quite serious and critical. Experiments on conventional transfer molding were conducted to evaluate wire span threshold with 23.0 μm diameter gold wire. The results show that the wire span threshold is about 4.1 mm, which is much shorter than the wire span threshold of over 5.0 mm for wire with 25.4 μm diameter. Finite element analysis shows there is a significant difference in the wire deflection between 23.0 μm gold wire and 25.4 μm gold wire diameter under the same action of mold flow. A novel encapsulation method is introduced using non-sweep solution. The wire span could be extended to over 5.0 mm with wire sweep less than 1%. Reliability tests conducted showed that all the units passed 1000 temperature cycles (−55 to 125 °C) with JEDEC moisture sensitivity level 2a (60 °C/60% relative humidity for 120 h) and 3 times reflow (peak temperature at 220–225 °C). It is believed that this solution could efficiently overcome the risk of wire short issues and improve the yield of ultra fine pitch wire bonds in high-volume production.     

A new method to estimate parameters of linear compartmental models using artificial neural networks

At present, the preferred tool for parameter estimation in compartmental analysis is an iterative procedure; weighted nonlinear regression. For a large number of applications, observed data can be fitted to sums of exponentials whose parameters are directly related to the rate constants/coefficients of the compartmental models. Since weighted nonlinear regression often has to be repeated for many different data sets, the process of fitting data from compartmental systems can be very time consuming. Furthermore the minimization routine often converges to a local (as opposed to global) minimum. In this paper, we examine the possibility of using artificial neural networks instead of weighted nonlinear regression in order to estimate model parameters. We train simple feed-forward neural networks to produce as outputs the parameter values of a given model when kinetic data are fed to the networks' input layer. The artificial neural networks produce unbiased estimates and are orders of magnitude faster than regression algorithms. At noise levels typical of many real applications, the neural networks are found to produce lower variance estimates than weighted nonlinear regression in the estimation of parameters from mono- and biexponential models. These results are primarily due to the inability of weighted nonlinear regression to converge. These results establish that artificial neural networks are powerful tools for estimating parameters for simple compartmental models.     

川东地区井下套管损坏的原因分析及对策

本文就近年来在钻井施工中遇到的４口井发套管损坏的问题进行了原因分析，提出了防止井下套管损坏对策。     

Fueguian cranial morphology: the adaptation to a cold, harsh environment

Craniometric data from the three extinct tribes that inhabited Tierra del Fuego (Selk'nam, Yámana, and Kawéskar) were gathered following Howell's measurement technique. We studied 180 skulls preserved at thirteen different institutions. Analysis of variance (ANOVA) between groups showed that morphological similarities among Fueguian groups are far more important than some differences between marine (Yámana and Kawéskar) and terrestrial (Selk'nam) groups. A principal component analysis (PCA) generated from the correlation matrix shows that Fueguians fall as outliers with respect to the typical Mongoloid morphology. In addition, a UPGMA tree generated from a squared Euclidean distance matrix indicates that Fueguian groups have a morphological pattern that is very distinct from that of other present-day Amerindian groups, with the exception of the Eskimos. One of the variables that contributes substantially to the differentiation of Eskimos and Fueguians is the nasal height. This suggests that nasal morphology in both groups could be a response to adaptive pressures related to the cold environment. However, other morphological particularities of Fueguian skulls, such as craniofacial robustness and variables of craniofacial width, can be attributed to a large masticatory stress. As a whole, the morphological features of Fueguian groups can be regarded as a general adaptive response to a very harsh environment, along with the retention of some plesiomorphic features. Assuming that the initial entry in Tierra de Fuego took place around 10,000 years BP, before the disappearance of the last land bridges in the Magellan Straits, then this adaptation might have arisen in a relatively short period, hastened by the extreme environmental conditions.     

The effects of limbic lesions on locomotion and stereotypy elicited by dopamine agonists in the rat

The purpose of this experiment was to investigate the functional contributions of various limbic structures to locomotion and stereotypy induced by dopaminergic drugs. Female rats were randomly assigned to one of 5 groups (n = 10-14 rats/group) that received either a lesion of the hippocampus (colchicine + kainic acid), basolateral amygdala (quinolinic acid), frontal cortex (aspiration), nucleus accumbens (ibotenic acid), or served as unoperated controls. Beginning at least 2 weeks following surgery locomotion (measured as photocell beam breaks) elicited by D-amphetamine (0.0, 0.32, 1.0 and 3.2 mg/kg), SKF 82958 (0.0, 0.04, 0.08 and 0.16 mg/kg) or quinpirole (0.0, 0.25, 0.1 and 0.5 mg/kg) was determined. In agreement with previous results rats with hippocampal lesions were hyperactive in response to amphetamine. In comparison to these changes in drug-induced locomotion, lesions of the basolateral amygdala, and frontal cortex had only minor effects on drug-induced locomotion. Lesions of the nucleus accumbens produced consistent hyperactivity that was suppressed by doses of amphetamine or quinpirole that elicited behavioral stereotypy. These results provide evidence suggesting that, in comparison to other limbic structures that have substantial inputs to the nucleus accumbens, the hippocampus play a relatively prominent role in the modulation of drug-induced locomotion.     

Response of Plasmodium falciparum to chloroquine and Fansidar in vivo and chloroquine and amodiaquine in vitro in Uganda

The response of P. falciparum to chloroquine and pyrimethamine-sulfadoxine in vivo and chloroquine and amodiaquine in vitro was investigated in parasitaemic school children from six locations. Mean parasite sensitivity to chloroquine at day 7 was 74% (range 61-97) with parasite clearance rates between 2-3 days and complete defervescence in 85% of febrile children. Sensitivity declined in the four sites followed up to day 14 to 45% (range 37-53). Parasites were significantly more sensitive to pyrimethamine/sulfadoxine at 5/6 sites (100% day 7) but 5% of subjects became parasitaemic by day 14. In vitro isolates were significantly less sensitive to chloroquine than to amodiaquine with a mean 99% effective concentration of 348 mumol/L compared to 6.44 mumol/L. Clearly the role of chloroquine as the primary therapy for uncomplicated P. falciparum malaria should be reconsidered especially in the light of increasing disease severity and resurgence. Amodiaquine may be suitable alternative with pyrimethamine/sulfadoxine as second line and for more severe malaria prior to referral. The cost of alternative antimalarials and the dynamic and deteriorating pattern of resistance are powerful arguments for more objective slide diagnosis to minimise drug pressure and a regular drug sensitivity surveillance system. We believe that the latter should concentrate on measuring clinical drug efficacy in symptomatic outpatients rather than in asymptomatic children while the former needs more pragmatic and economical strategies possibly centred on seasonality and risk.