用于微阵列数据癌症分类的演化硬件多分类器

针对单分类器识别率低、稳定性差的问题,提出了一种用于微阵列数据分类的演化硬件多分类器选择性集成方法.首先把经过预处理的原始训练集随机划分为训练集和验证集;然后通过对训练集的学习获得基于演化硬件的基分类器;再用验证集评价基分类器的性能,选择其中一部分较好的基分类器组成最终的分类系统;最后用独立的测试集验证系统的性能.试验结果表明,对急性白血病和结肠癌数据集的识别率分别为95.42％、88.33％,与其他的模式识别方法具有可比性;同时在识别率相当的情况下,该方法的硬件代价远低于全集成的演化硬件多分类器.     

Initial corrosion observed on the atomic scale

Corrosion destroys more than three per cent of the world's GDP. Recently, the electrochemical decomposition of metal alloys has been more productively harnessed to produce porous materials with diverse technological potential. High-resolution insight into structure formation during electrocorrosion is a prerequisite for an atomistic understanding and control of such electrochemical surface processes. Here we report atomic-scale observations of the initial stages of corrosion of a Cu3Au111 single crystal alloy within a sulphuric acid solution. We monitor, by in situ X-ray diffraction with picometre-scale resolution, the structure and chemical composition of the electrolyte/alloy interface as the material decomposes. We reveal the microscopic structural changes associated with a general passivation phenomenon of which the origin has been hitherto unclear. We observe the formation of a gold-enriched single-crystal layer that is two to three monolayers thick, and has an unexpected inverted (CBA-) stacking sequence. At higher potentials, we find that this protective passivation layer dewets and pure gold islands are formed; such structures form the templates for the growth of nanoporous metals. Our experiments are carried out on a model single-crystal system. However, the insights should equally apply within a crystalline grain of an associated polycrystalline electrode fabricated from many other alloys exhibiting a large difference in the standard potential of their constituents, such as stainless steel (see ref. 5 for example) or alloys used for marine applications, such as CuZn or CuAl.     

Mammalian cochlear supporting cells can divide and trans-differentiate into hair cells

Sensory hair cells of the mammalian organ of Corti in the inner ear do not regenerate when lost as a consequence of injury, disease, or age-related deafness. This contrasts with other vertebrates such as birds, where the death of hair cells causes surrounding supporting cells to re-enter the cell cycle and give rise to both new hair cells and supporting cells. It is not clear whether the lack of mammalian hair cell regeneration is due to an intrinsic inability of supporting cells to divide and differentiate or to an absence or blockade of regenerative signals. Here we show that post-mitotic supporting cells purified from the postnatal mouse cochlea retain the ability to divide and trans-differentiate into new hair cells in culture. Furthermore, we show that age-dependent changes in supporting cell proliferative capacity are due in part to changes in the ability to downregulate the cyclin-dependent kinase inhibitor p27(Kip1) (also known as Cdkn1b). These results indicate that postnatal mammalian supporting cells are potential targets for therapeutic manipulation.     

Design of child’s safety management system based on ontology using context information of home network

This paper presents a system to alert of dangerous a child situation of a child by applying context information collected from a home network to ontology that is capable of inference. Radio frequency Identification (RFID) and sensors were used for the configuration of a home network, to obtain the raw data to convert into context information. To express the ontology, web ontology language (OWL) was used to provide the inference of context information. Then, simple object access protocol (SOAP) messages were used to notify of the dangerous situations that a child may be involved in via mobile devices. The proposed system consists of Context Manager, Service Manager, and Notification Manager. The child's safety management system can proactively detect the context data of a child on the basis of context awareness. In the experiment, the Jena 2.0 by ontology reasoner and the OSGi(Open Service Gateway initiative) Gateway developed using open source software Knopflerfish 1.3.3 were used to implement the service frame work.     

Active Cooling e-Textiles for Smart Clothing

Cooling function is definitely one of the most desirable attributes of clothing. In spite of the recent progress on phase changing material （PCM） research, the final products with sufficient amount of cooling capability have not yet to be developed in market. A new concept of cooling fabrics has been proposed by applying ＂Peltier effect＂ to textile materials. It occurs whenever electrical current flows through two dissimilar conductors t depending on the direction of current flow, the junction of the two conductors will either absorb or release heat. This effect has been tested on P-type and N-type conducting polymers. A P-type conductive polypyrrole coated fabric was synthesized by in- situ polymerization on plain weave PET to make conductive fabrics. And an N.typo electrically conductive material was synthesized by treatment of MWNT and polyethyleneimine （PEI）. A noticeable amount of temperature difference has been found in the fabric.     

Identifying the time‐effect factors of multiple time series

The Peña–Box model is considered for finding the time‐effect factors of a multiple time series. This paper first establishes the connection between the Peña–Box model and the vector ARMA model. According to the Peña–Box model, some series can be ignored while modelling the vector ARMA model. A consistent estimator is then proposed to identify the model for nonlinear and nonstationary time series. Finally, the finite‐sample behaviour of the estimator is illustrated via simulations. Copyright © 2005 John Wiley & Sons, Ltd.     

A Bayesian threshold nonlinearity test for financial time series

We propose in this paper a threshold nonlinearity test for financial time series. Our approach adopts reversible‐jump Markov chain Monte Carlo methods to calculate the posterior probabilities of two competitive models, namely GARCH and threshold GARCH models. Posterior evidence favouring the threshold GARCH model indicates threshold nonlinearity or volatility asymmetry. Simulation experiments demonstrate that our method works very well in distinguishing GARCH and threshold GARCH models. Sensitivity analysis shows that our method is robust to misspecification in error distribution. In the application to 10 market indexes, clear evidence of threshold nonlinearity is discovered and thus supporting volatility asymmetry. Copyright © 2005 John Wiley & Sons, Ltd.     

HARD DISK DRIVES CONTROL IN MOBILE APPLICATIONS

In this paper, control design is investigated for hard disk drives in mobile applications with unknown external arbitrarily fast time-varying disturbances. The disturbances can be estimated with exponential accuracy using the proposed disturbance observer based on a series of integral filters. The position error signal will converge to zero with the proposed control technique for systems subjected to the unknown disturbances. The effectiveness of the proposed method is demonstrated by extensive simulation studies.     

Technological forecasting with nonlinear models

The S-shaped growth curves such as Gompertz, logistic, normal and Weibuli are widely used for forecasting technological substitutions. A family of data-based transformed (DBT) models, which are linear in the regression parameters, including the above-mentioned four models as special cases has been shown to be quite useful for short-term forecasts. This paper explores modeling the technology penetration data directly with assumed S-shaped growth curves. The resulting models, which are nonlinear in the regression parameters, also incorporate proper dependence structure and power transformation. It appears that the nonlinear modeling is a viable alternative to the DBT and other conventional forecasting models in forecasting technological substitutions. Hence, an appropriate strategy is to consider the nonlinear modeling approaches as possible alternatives and use the data at hand to select, via pseudo-cross-validation, the best model for forecasting purposes.