Showing papers by "Nanjing University published in 2005"

Tri-training: exploiting unlabeled data using three classifiers

Abstract: In many practical data mining applications, such as Web page classification, unlabeled training examples are readily available, but labeled ones are fairly expensive to obtain. Therefore, semi-supervised learning algorithms such as co-training have attracted much attention. In this paper, a new co-training style semi-supervised learning algorithm, named tri-training, is proposed. This algorithm generates three classifiers from the original labeled example set. These classifiers are then refined using unlabeled examples in the tri-training process. In detail, in each round of tri-training, an unlabeled example is labeled for a classifier if the other two classifiers agree on the labeling, under certain conditions. Since tri-training neither requires the instance space to be described with sufficient and redundant views nor does it put any constraints on the supervised learning algorithm, its applicability is broader than that of previous co-training style algorithms. Experiments on UCI data sets and application to the Web page classification task indicate that tri-training can effectively exploit unlabeled data to enhance the learning performance.

EECS: an energy efficient clustering scheme in wireless sensor networks

Abstract: Data gathering is a common but critical operation in many applications of wireless sensor networks. Innovative techniques that improve energy efficiency to prolong the network lifetime are highly required. Clustering is an effective topology control approach in wireless sensor networks, which can increase network scalability and lifetime. In this paper, we propose a novel clustering schema EECS for wireless sensor networks, which better suits the periodical data gathering applications. Our approach elects cluster heads with more residual energy through local radio communication while achieving well cluster head distribution; further more it introduces a novel method to balance the load among the cluster heads. Simulation results show that EECS outperforms LEACH significantly with prolonging the network lifetime over 35%.

An energy-efficient unequal clustering mechanism for wireless sensor networks

Abstract: Clustering provides an effective way for prolonging the lifetime of a wireless sensor network. Current clustering algorithms usually utilize two techniques, selecting cluster heads with more residual energy and rotating cluster heads periodically, to distribute the energy consumption among nodes in each cluster and extend the network lifetime. However, they rarely consider the hot spots problem in multihop wireless sensor networks. When cluster heads cooperate with each other to forward their data to the base station, the cluster heads closer to the base station are burdened with heavy relay traffic and tend to die early, leaving areas of the network uncovered and causing network partition. To address the problem, we propose an energy-efficient unequal clustering (EEUC) mechanism for periodical data gathering in wireless sensor networks. It partitions the nodes into clusters of unequal size, and clusters closer to the base station have smaller sizes than those farther away from the base station. Thus cluster heads closer to the base station can preserve some energy for the inter-cluster data forwarding. We also propose an energy-aware multihop routing protocol for the inter-cluster communication. Simulation results show that our unequal clustering mechanism balances the energy consumption well among all sensor nodes and achieves an obvious improvement on the network lifetime

The mitochondrial uncoupling-protein homologues.

Abstract: Uncoupling protein(UCP)1 is an integral membrane protein that is located in the mitochondrial inner membrane of brown adipocytes. Its physiological role is to mediate a regulated, thermogenic proton leak. UCP2 and UCP3 are recently identified UCP1 homologues. They also mediate regulated proton leak, and might function to control the production of superoxide and other downstream reactive oxygen species. However, their role in normal physiology remains unknown. Recent studies have shown that UCP2 has an important part in the pathogenesis of type-2 diabetes. The obscure roles of the UCP homologues in normal physiology, together with their emerging role in pathophysiology, provide exciting potential for further investigation.

A k-nearest neighbor based algorithm for multi-label classification

Abstract: In multi-label learning, each instance in the training set is associated with a set of labels, and the task is to output a label set whose size is unknown a priori for each unseen instance. In this paper, a multi-label lazy learning approach named ML-kNN is presented, which is derived from the traditional k-nearest neighbor (kNN) algorithm. In detail, for each new instance, its k-nearest neighbors are firstly identified. After that, according to the label sets of these neighboring instances, maximum a posteriori (MAP) principle is utilized to determine the label set for the new instance. Experiments on a real-world multi-label bioinformatic data show that ML-kNN is highly comparable to existing multi-label learning algorithms.

Highly Ordered Platinum‐Nanotubule Arrays for Amperometric Glucose Sensing

Abstract: Direct glucose sensing on highly ordered platinum-nanotubule array electrodes (NTAEs) is systematically investigated. The NTAEs are fabricated by electrochemical deposition of platinum in a 3-aminopropyltrimethoxysilane-modified anodic alumina membrane. Their structures and morphologies are then characterized using X-ray diffraction and scanning electron microscopy, respectively. Electrochemical results show that NTAEs with different real surface areas could be achieved by controlling the deposition time or by using anodic alumina membranes with different pore size. Electrochemical responses of the as-synthesized NTAEs to glucose in a solutions of either 0.5 M H2SO4, or phosphate-buffered saline (PBS, pH 7.4) containing 0.1 M KCl are discussed. Based on the different electrochemical reaction mechanisms of glucose and interferents such as p-acetamedophenol and ascorbic acid, their high roughness factor makes NTAEs sensitive, selective, and stable enough to be a kind of biosensor for the non-enzymatic detection of glucose. Such a glucose sensor allows the determination of glucose in the linear range 2–14 mM, with a sensitivity of 0.1 μA cm–2 mM–1 (correlation coefficient 0.999), and a detection limit of 1.0 μM glucose, with neglectable interference from physiological levels of 0.1 mM p-acetamedophenol, 0.1 mM ascorbic acid, and 0.02 mM uric acid.

Supervised nonlinear dimensionality reduction for visualization and classification

Abstract: When performing visualization and classification, people often confront the problem of dimensionality reduction. Isomap is one of the most promising nonlinear dimensionality reduction techniques. However, when Isomap is applied to real-world data, it shows some limitations, such as being sensitive to noise. In this paper, an improved version of Isomap, namely S-Isomap, is proposed. S-Isomap utilizes class information to guide the procedure of nonlinear dimensionality reduction. Such a kind of procedure is called supervised nonlinear dimensionality reduction. In S-Isomap, the neighborhood graph of the input data is constructed according to a certain kind of dissimilarity between data points, which is specially designed to integrate the class information. The dissimilarity has several good properties which help to discover the true neighborhood of the data and, thus, makes S-Isomap a robust technique for both visualization and classification, especially for real-world problems. In the visualization experiments, S-Isomap is compared with Isomap, LLE, and WeightedIso. The results show that S-Isomap performs the best. In the classification experiments, S-Isomap is used as a preprocess of classification and compared with Isomap, WeightedIso, as well as some other well-established classification methods, including the K-nearest neighbor classifier, BP neural network, J4.8 decision tree, and SVM. The results reveal that S-Isomap excels compared to Isomap and WeightedIso in classification, and it is highly competitive with those well-known classification methods.

Synthesis of Titanium Dioxide Nanoparticles with Mesoporous Anatase Wall and High Photocatalytic Activity

Abstract: Mesoporous titanium dioxide nanosized powder with high specific surface area and anatase wall was synthesized via hydrothermal process by using cetyltrimethylammonium bromide (CTAB) as surfactant-directing agent and pore-forming agent. The resulting materials were characterized by XRD, nitrogen adsorption, FESEM, TEM, and FT-IR spectroscopy. The as-synthesized mesoporous TiO2 nanoparticles have mean diameter of 17.6 nm with mean pore size of 2.1 nm. The specific surface area of the as-synthesized mesoporous nanosized TiO2 exceeded 430 m2/g and that of the samples after calcination at 600 degrees C still have 221.9 m2/g. The mesoporous TiO2 nanoparticles show significant activities on the oxidation of Rhodamine B (RB). The large surface area, small crystalline size, and well-crystallized anatase mesostructure can explain the high photocatalytic activity of mesoporous TiO2 nanoparticles calcined at 400 degrees C.

Ab initio investigations of optical properties of the high-pressure phases of ZnO

Abstract: We present a detailed investigation on optical properties of high-pressure phase ZnO in $B1$ (NaCl) and $B2$ (CsCl) structures, including dielectric function, refractive index, absorption, and electron energy-loss spectrum. Theoretical calculations are performed using the ab initio pseudopotential density functional method, in which we employ the Perdew-Burke-Eruzerhof form of the generalized gradient approximation available in the CASTEP code together with plane wave basis sets for expanding the periodic electron density. Both structures are optimized under the respective structural phase transition pressures; for the $B1$ structure it is $9\phantom{\rule{0.3em}{0ex}}\mathrm{GPa}$, which has been verified to be in agreement with theory [Jaffe et al., Phys. Rev. B 62, 1660 (2000)] and experiment [Desgreniers, Phys. Rev. B 58, 14102 (1998)], while for the $B2$ structure a transition pressure of $256\phantom{\rule{0.3em}{0ex}}\mathrm{GPa}$ is predicted in theory [Jaffe et al., Phys. Rev. B 62, 1660 (2000)]. We find that their electronic structures and optical properties under high pressure are quite different from those under ambient pressure.

1D helix, 2D brick-wall and herringbone, and 3D interpenetration d10 metal-organic framework structures assembled from pyridine-2,6-dicarboxylic acid N-oxide.

Abstract: Five novel interesting d10 metal coordination polymers, [Zn(PDCO)(H2O)2]n (PDCO = pyridine-2,6-dicarboxylic acid N-oxide) (1), [Zn2(PDCO)2(4,4‘-bpy)2(H2O)2·3H2O]n (bpy = bipyridine) (2), [Zn(PDCO)(bix)]n (bix = 1,4-bis(imidazol-1-ylmethyl)benzene) (3), [Zn(PDCO)(bbi)·0.5H2O]n (bbi = 1,1‘-(1,4-butanediyl)bis(imidazole)) (4), and [Cd(PDCO)(bix)1.5·1.5H2O]n (5), have been synthesized under hydrothermal conditions and structurally characterized. Polymer 1 possesses a one-dimensional (1D) helical chainlike structure with 41 helices running along the c-axis with a pitch of 10.090 A. Polymer 2 has an infinite chiral two-dimensional (2D) brick-wall-like layer structure in the ac plane built from achiral components, while both 3 and 4 exhibit an infinite 2D herringbone architecture, respectively extended in the ac and ab plane. Polymer 5 features a most remarkable and unique three-dimensional (3D) porous framework with 2-fold interpenetration related by symmetry, which contains channels in the b and c directions, ...

Experimental evidence for the quantum confinement effect in 3C-SiC nanocrystallites.

Abstract: Using electrochemical etching of a polycrystalline 3$C$-SiC target and subsequent ultrasonic treatment in water solution, we have fabricated suspensions of 3$C$-SiC nanocrystallites that luminesce. Transmission electron microscope observations show that the 3$C$-SiC nanocrystallites, which uniformly disperse in water, have sizes in the range of 1\char21{}6 nm. Photoluminescence and photoluminescence excitation spectral examinations show clear evidence for the quantum confinement of 3$C$-SiC nanocrystallites with the emission band maximum ranging from 440 to 560 nm. Tunable, composite polystyrene/SiC film can be made by adding polystyrene to a toluene suspension of the 3$C$-SiC nanocrystallites and then coating the resulting solution onto a Si wafer.

Facile Fabrication of Functional Polypyrrole Nanotubes via a Reactive Self-Degraded Template

Abstract: Summary: Polypyrrole nanotubes with high electric conductivity and azo function have been fabricated in high yield via an in-situ polymerization. During the process fibrillar complex of FeCl3 and methyl orange (MO), acting as a reactive self-degraded template, directed the growth of polypyrrole on its surface and promoted the assembly into hollow nanotubular structures. TEM image of uncompleted PPy nanotubes synthesized in MO solutions after reaction for 40 min.

Fault estimation and accommodation for linear MIMO discrete-time systems

Abstract: In this brief, a methodology for detection and accommodation of actuator faults for a class of multi-input-multi-output (MIMO) stochastic systems is presented. First, a new real-time fault estimation module that estimates the actuator effectiveness is developed. The actuator fault diagnosis is based on the estimation of the state vector. Under some conditions, the stochastic system is transformed into two separate subsystems. One of them is not affected by actuator faults, so a reduced order Kalman filter can be used to estimate its states. The other, whose states are measurable, is affected by the faults. Then, the output of the nominal controller is reconfigured to compensate for the loss of actuator effectiveness in the system. Simulation results of a helicopter in vertical plane is presented to demonstrate the performance of the proposed fault-tolerant control scheme.

Recognizing partially occluded, expression variant faces from single training image per person with SOM and soft k-NN ensemble

Abstract: Most classical template-based frontal face recognition techniques assume that multiple images per person are available for training, while in many real-world applications only one training image per person is available and the test images may be partially occluded or may vary in expressions. This paper addresses those problems by extending a previous local probabilistic approach presented by Martinez, using the self-organizing map (SOM) instead of a mixture of Gaussians to learn the subspace that represented each individual. Based on the localization of the training images, two strategies of learning the SOM topological space are proposed, namely to train a single SOM map for all the samples and to train a separate SOM map for each class, respectively. A soft k nearest neighbor (soft k-NN) ensemble method, which can effectively exploit the outputs of the SOM topological space, is also proposed to identify the unlabeled subjects. Experiments show that the proposed method exhibits high robust performance against the partial occlusions and variant expressions.

Semi-supervised regression with co-training

Abstract: In many practical machine learning and data mining applications, unlabeled training examples are readily available but labeled ones are fairly expensive to obtain. Therefore, semi-supervised learning algorithms such as co-training have attracted much attention. Previous research mainly focuses on semi-supervised classification. In this paper, a co-training style semi-supervised regression algorithm, i.e. COREG, is proposed. This algorithm uses two k-nearest neighbor regressors with different distance metrics, each of which labels the unlabeled data for the other regressor where the labeling confidence is estimated through consulting the influence of the labeling of unlabeled examples on the labeled ones. Experiments show that COREG can effectively exploit unlabeled data to improve regression estimates.

Octacyanometallate-Based Single-Molecule Magnets: CoII9MV6 (M = W, Mo)

Abstract: Two octacyanometallate-based clusters, {CoII9[WV(CN)8]6·(CH3OH)24}·19H2O (1) and {CoII9[MoV(CN)8]6·(CH3OH)24}·4CH3OH·16H2O (2), have been synthesized. Both complexes show the single-molecule magnet behavior.

Nonenzymatic Glucose Detection by Using a Three‐Dimensionally Ordered, Macroporous Platinum Template

Abstract: A three-dimensionally ordered, macroporous, inverse-opal platinum film was synthesized electrochemically by the inverted colloidal-crystal template technique. The inverse-opal film that contains platinum nanoparticles showed improved electrocatalytic activity toward glucose oxidation with respect to the directly deposited platinum; this improvement is due to the interconnected porous structure and the greatly enhanced effective surface area. In addition, the inverse-opal Pt-film electrode responds more sensitively to glucose than to common interfering species of ascorbic acid, uric acid, and p-acetamidophenol due to their different electrochemical reaction mechanisms. Results showed that the ordered macroporous materials with enhanced selectivity and sensitivity are promising for fabrication of nonenzymatic glucose biosensors.

A full view of eit waves

Abstract: Early observations by the EUV Imaging Telescope (EIT) on the Solar and Heliospheric Observatory indicated that propagating diffuse wave fronts, now conventionally referred to as "EIT waves," can often be seen on the solar disk with a propagation velocity several times smaller than that of Hα Moreton waves. They are almost always associated with coronal mass ejections. We have previously confirmed the existence of such a wave phenomenon with numerical simulations, which indicate that there does exist a slower moving "wave" much behind the coronal counterpart of the Hα Moreton wave. Further observations have disclosed many new features of the EIT waves: the waves stop near the separatrix between active regions, sometimes they experience acceleration from the active region to the quiet region, and so on. Here we report on MHD simulations performed to demonstrate how the typical features of EIT waves can all be accounted for within our theoretical model, in which the EIT waves are thought to be formed by successive stretching or opening of closed field lines driven by an erupting flux rope. The relationship between EIT waves, Hα Moreton waves, and type II radio bursts is discussed, with an emphasis on reconciling the discrepancies among different views of the "EIT wave" phenomenon.