Showing papers by "Nanjing University published in 2007"

Top 10 algorithms in data mining

Abstract: This paper presents the top 10 data mining algorithms identified by the IEEE International Conference on Data Mining (ICDM) in December 2006: C4.5, k-Means, SVM, Apriori, EM, PageRank, AdaBoost, kNN, Naive Bayes, and CART. These top 10 algorithms are among the most influential data mining algorithms in the research community. With each algorithm, we provide a description of the algorithm, discuss the impact of the algorithm, and review current and further research on the algorithm. These 10 algorithms cover classification, clustering, statistical learning, association analysis, and link mining, which are all among the most important topics in data mining research and development.

Low temperature preparation and visible light photocatalytic activity of mesoporous carbon-doped crystalline TiO2

Abstract: A visible-light-active TiO2 photocatalyst was prepared through carbon doping by using glucose as carbon source. Different from the previous carbon-doped TiO2 prepared at high temperature, our preparation was performed by a hydrothermal method at temperature as low as 160 °C. The resulting photocatalyst was characterized by XRD, XPS, TEM, nitrogen adsorption, and UV–vis diffuse reflectance spectroscopy. The characterizations found that the photocatalyst possessed a homogeneous pore diameter about 8 nm and a high surface area of 126 m2/g. Comparing to undoped TiO2, the carbon-doped TiO2 showed obvious absorption in the 400–450 nm range with a red shift in the band gap transition. It was found that the resulting carbon-doped TiO2 exhibits significantly higher photocatalytic activity than the undoped counterpart and Degussa P25 on the degradation of rhodamine B (RhB) in water under visible light irradiation (λ > 420 nm). This method can be easily scaled up for industrial production of visible-light driven photocatalyst for pollutants removal because of its convenience and energy-saving.

A Leaf Recognition Algorithm for Plant Classification Using Probabilistic Neural Network

Abstract: In this paper, we employ probabilistic neural network (PNN) with image and data processing techniques to implement a general purpose automated leaf recognition for plant classification. 12 leaf features are extracted and orthogonalized into 5 principal variables which consist the input vector of the PNN. The PNN is trained by 1800 leaves to classify 32 kinds of plants with an accuracy greater than 90%. Compared with other approaches, our algorithm is an accurate artificial intelligence approach which is fast in execution and easy in implementation.

Visible-Light Photocatalytic Properties of Weak Magnetic BiFeO3 Nanoparticles

Abstract: Polycrystalline BiFeO3 nanoparticles (size 80-120 nm) are prepared by a simple sol-gel technique. Such nanoparticles are very efficient for photocatalytic decomposition of organic contaminants under irradiation from ultraviolet to visible frequencies. The BiFeO3 nanoparticles also demonstrate weak ferromagnetism of about 0.06 mu(B)/Fe at room temperature, in good agreement with theoretical calculations.

Adsorption of polar and nonpolar organic chemicals to carbon nanotubes.

Abstract: Understanding adsorptive interactions between organic contaminants and carbon nanotubes is critical to both the environmental application of carbon nanotubes as special adsorbents and the assessment of the potential impact of carbon nanotubes on the fate and transport of organic contaminants in the environment. The adsorption of organic compounds with varied physical−chemical properties (hydrophobicity, polarity, electron polarizability, and size) to one single-walled carbon nanotube (SWNT) and two multiwalled carbon nanotubes (MWNTs) was evaluated. For a given carbon nanotube, the adsorption affinity correlated poorly with hydrophobicity but increased in the order of nonpolar aliphatic < nonpolar aromatics < nitroaromatics, and within the group of nitroaromatics, the adsorption affinity increased with the number of nitro-functional groups. We propose that the strong adsorptive interaction between carbon nanotubes and nitroaromatics was due to the π−π electron-donor–acceptor (EDA) interaction between nitr...

Glucose sensing by POMC neurons regulates glucose homeostasis and is impaired in obesity

Abstract: A subset of neurons in the brain, known as 'glucose-excited' neurons, depolarize and increase their firing rate in response to increases in extracellular glucose. Similar to insulin secretion by pancreatic beta-cells, glucose excitation of neurons is driven by ATP-mediated closure of ATP-sensitive potassium (K(ATP)) channels. Although beta-cell-like glucose sensing in neurons is well established, its physiological relevance and contribution to disease states such as type 2 diabetes remain unknown. To address these issues, we disrupted glucose sensing in glucose-excited pro-opiomelanocortin (POMC) neurons via transgenic expression of a mutant Kir6.2 subunit (encoded by the Kcnj11 gene) that prevents ATP-mediated closure of K(ATP) channels. Here we show that this genetic manipulation impaired the whole-body response to a systemic glucose load, demonstrating a role for glucose sensing by POMC neurons in the overall physiological control of blood glucose. We also found that glucose sensing by POMC neurons became defective in obese mice on a high-fat diet, suggesting that loss of glucose sensing by neurons has a role in the development of type 2 diabetes. The mechanism for obesity-induced loss of glucose sensing in POMC neurons involves uncoupling protein 2 (UCP2), a mitochondrial protein that impairs glucose-stimulated ATP production. UCP2 negatively regulates glucose sensing in POMC neurons. We found that genetic deletion of Ucp2 prevents obesity-induced loss of glucose sensing, and that acute pharmacological inhibition of UCP2 reverses loss of glucose sensing. We conclude that obesity-induced, UCP2-mediated loss of glucose sensing in glucose-excited neurons might have a pathogenic role in the development of type 2 diabetes.

A functional polymorphism in the 5′ UTR of GDF5 is associated with susceptibility to osteoarthritis

Abstract: Osteoarthritis (MIM 165720), characterized by degeneration of articular cartilage, is the most common form of human arthritis and a major concern for aging societies worldwide. Epidemiological and genetic studies have shown that osteoarthritis is a polygenic disease. Here, we report that the gene encoding growth differentiation factor 5 (GDF5) is associated with osteoarthritis in Asian populations. A SNP in the 5' UTR of GDF5 (+104T/C; rs143383) showed significant association (P = 1.8 x 10(-13)) with hip osteoarthritis in two independent Japanese populations. This association was replicated for knee osteoarthritis in Japanese (P = 0.0021) and Han Chinese (P = 0.00028) populations. This SNP, located in the GDF5 core promoter, exerts allelic differences on transcriptional activity in chondrogenic cells, with the susceptibility allele showing reduced activity. Our findings implicate GDF5 as a susceptibility gene for osteoarthritis and suggest that decreased GDF5 expression is involved in the pathogenesis of osteoarthritis.

Generation of arbitrary vector beams with a spatial light modulator and a common path interferometric arrangement

Abstract: We describe a convenient approach for generating arbitrary vector beams in a 4-f system with a spatial light modulator (SLM) and a common path interferometric arrangement. A computer-generated hologram is introduced onto SLM for performing the beam conversion. Optical realization of a variety of polarization configurations confirms the reliability and flexibility of our method.

Biomedical and clinical applications of immunoassays and immunosensors for tumor markers

Abstract: Cancer is one of the most threatening diseases for human beings. The detection of tumor biomarkers, including serum tumor markers and potential prognostic factors for tumors, plays an important role in clinical diagnoses and evaluation of treatment for patients with certain tumor-associated diseases. Many immunoassay methods and immunosensors are being developed for detection of tumor-related biomarkers. This review summarizes the state of immunoassay technologies for tumor diagnostics. It also focuses on a growing number of applications and progress in immunosensors for tumor markers. It covers the basic principles and biomedical and clinical applications of immunosensors, and indicates the future prospects in this field.

A gold nanoparticle-based aptamer target binding readout for ATP assay

Abstract: A gold nanoparticle-based detection strategy for ATP assays. In the absence of ATP, gold nanoparticles are not stabilized by the rigid duplex, thus they are readily aggregated by salt (solution displaying blue colors); in contrast induces a duplex-to-aptamer structural switching, liberating a random coil-like ssDNA. Gold nanoparticles are stabilized by the liberated ssDNA, showing high resistance to salt-induced aggregation (solution staying in red).

Improve Computer-Aided Diagnosis With Machine Learning Techniques Using Undiagnosed Samples

Abstract: In computer-aided diagnosis (CAD), machine learning techniques have been widely applied to learn a hypothesis from diagnosed samples to assist the medical experts in making a diagnosis. To learn a well-performed hypothesis, a large amount of diagnosed samples are required. Although the samples can be easily collected from routine medical examinations, it is usually impossible for medical experts to make a diagnosis for each of the collected samples. If a hypothesis could be learned in the presence of a large amount of undiagnosed samples, the heavy burden on the medical experts could be released. In this paper, a new semisupervised learning algorithm named Co-Forest is proposed. It extends the co-training paradigm by using a well-known ensemble method named Random Forest, which enables Co-Forest to estimate the labeling confidence of undiagnosed samples and easily produce the final hypothesis. Experiments on benchmark data sets verify the effectiveness of the proposed algorithm. Case studies on three medical data sets and a successful application to microcalcification detection for breast cancer diagnosis show that undiagnosed samples are helpful in building CAD systems, and Co-Forest is able to enhance the performance of the hypothesis that is learned on only a small amount of diagnosed samples by utilizing the available undiagnosed samples.

Semi-Supervised Dimensionality Reduction.

Abstract: Dimensionality reduction is among the keys in mining highdimensional data. This paper studies semi-supervised dimensionality reduction. In this setting, besides abundant unlabeled examples, domain knowledge in the form of pairwise constraints are available, which specifies whether a pair of instances belong to the same class (must-link constraints) or different classes (cannot-link constraints). We propose the SSDR algorithm, which can preserve the intrinsic structure of the unlabeled data as well as both the must-link and cannot-link constraints defined on the labeled examples in the projected low-dimensional space. The SSDR algorithm is efficient and has a closed form solution. Experiments on a broad range of data sets show that SSDR is superior to many established dimensionality reduction methods.

Syntheses, Structures, and Physical Properties of Three Novel Metal−Organic Frameworks Constructed from Aromatic Polycarboxylate Acids and Flexible Imidazole-Based Synthons

Abstract: Three novel interesting coordination polymers, [Cd3(SIP)2(bbi)5·3H2O]n (1), [Co3(SIP)2(bix)4(2H2O)·2H2O]n (2), and [Cd(2,5-pydc)2(bix)1.5·H2O]n (3) (SIP = 5-sulfoisophthalic acid monosodium salt; bbi = 1,1‘-(1,4-butanediyl)bis(imidazole); bix = 1,4-bis(imidazol-1-ylmethyl)-benzene; 2,5-pydc = pyridine-2,5-dicarboxylic acid), have been isolated under hydrothermal conditions and structurally characterized. Polymer 1 has a 3D complicated framework comprised of an infinite 1D ladder-like chain and 2D layer structure. Polymer 2 features a 3D hydrated framework with uncoordinated water molecules trapped in the pores. Polymer 3 is a 2D infinite layer framework, which is further interconnected by hydrogen-bond interactions to lead to a 3D supramolecular architecture. Compounds 1 and 3 exhibit medium-strong fluorescent emissions in the solid state at room temperature and could be significant in the field of photoactive materials.

Low-Luminosity Gamma-Ray Bursts as a Unique Population: Luminosity Function, Local Rate, and Beaming Factor

Abstract: Swift BAT has detected ~200 long-duration GRBs, with redshift measurements for ~50 of them. We derive the luminosity function (?HL) and the local event rate (?) of the conventional high-luminosity (HL) GRBs by using the z-known Swift GRBs. Our results are generally consistent with that derived from the CGRO BATSE data. However, the fact that Swift detected a low-luminosity (LL) GRB, GRB 060218, at z = 0.033 within ~2 years of operation, together with the previous detection of the nearby GRB 980425, suggests a much higher local rate for these LL-GRBs. We explore the possibility that LL-GRBs are a distinct GRB population from the HL-GRBs. We find that ? is ~325 Gpc-3 yr-1, which is much higher than ? (1.12 Gpc-3 yr-1). This rate is ~0.7% of the local Type Ib/c SNe. Our results, together with the finding that less than 10% of Type Ib/c SNe are associated with off-beam GRBs, suggest that LL-GRBs have a beaming factor typically less than 14, or a jet angle typically wider than 31?. The high local GRB rate, small beaming factor, and low-luminosity make the LL-GRBs distinct from the HL-GRBs. Although the current data could not fully rule out the possibility that both HL- and LL-GRBs are the same population, our results suggest that LL-GRBs are likely a unique GRB population and that the observed low-redshift GRB sample is dominated by the LL-GRBs.

CdS Nanocrystal-Based Electrochemiluminescence Biosensor for the Detection of Low-Density Lipoprotein by Increasing Sensitivity with Gold Nanoparticle Amplification

Abstract: Mercaptoacetic acid (RSH)-capped CdS nanocrystals (NCs) was demonstrated to be electrochemically reduced during potential scan and react with the coreactant S2O82- to generate strong electrochemiluminescence (ECL) in aqueous solution. Based on the ECL of CdS NCs, a novel label-free ECL biosensor for the detection of low-density lipoprotein (LDL) has been developed by using self-assembly and gold nanoparticle amplification techniques. The biosensor was prepared as follows: The gold nanoparticles were first assembled onto a cysteamine monolayer on the gold electrode surface. This gold nanoparticle-covered electrode was next treated with cysteine and then reacted with CdS NCs to afford a CdS NC-electrode. Finally, apoB-100 (ligand of LDL receptor) was covalently conjugated to the CdS NC-electrode. The modification procedure was characterized by cyclic voltammetry, electrochemical impedance spectroscopy, and atomic force microscopy, respectively. The resulting modified electrode was tested as ECL biosensor f...

Spartina invasion in China: implications for invasive species management and future research

Abstract: Summary Four species of Spartina (Spartina anglica, S. alterniflora, S. patens and S. cynosuroides) have been introduced to China, but currently only the first three are present and only the first two successfully reproduce on the Chinese coast. Spartina anglica and S. alterniflora were introduced to China from England in 1963 and from the United States in 1979, respectively. Today, S. alterniflora has expanded its coverage to more than 112 000 ha and S. anglica has declined to <50 ha. This is compared with only 260 ha of S. alterniflora and over 36 000 ha of S. anglica in 1985. The fates of Chinese Spartina, with dramatic expansion of S. alterniflora and significant decline of S. anglica, were different from those in other locations throughout the world. Factors affecting the growth of the two naturalized Spartina species in China include differences in artificial plantation strategy, impacts of tideland reclamation, species competition ability and genetic diversity. Several methods for Spartina control in China, such as harvesting, herbicide application and freshwater irrigation, have been developed, but more research is needed to verify their effectiveness.

Superhydrophobicity of 3D Porous Copper Films Prepared Using the Hydrogen Bubble Dynamic Template

Abstract: We have successfully sculptured a variety of copper films with open interconnected macroporous walls and nanoparticles using hydrogen bubbles as the dynamic template. In this process, the hydrogen ...