Showing papers by "National University of Singapore published in 2003"
TL;DR: In this article, a comprehensive review is presented on the researches and developments related to electrospun polymer nanofibers including processing, structure and property characterization, applications, and modeling and simulations.
6,987 citations
TL;DR: In this paper, the authors examined the impact of ownership structure and board composition on voluntary disclosure and found that lower managerial ownership and significant government ownership are associated with increased disclosure, however, blockholder ownership is not related to disclosure.
1,898 citations
TL;DR: In the Fall of 2000, a database of more than 40,000 facial images of 68 people was collected using the Carnegie Mellon University 3D Room to imaged each person across 13 different poses, under 43 different illumination conditions, and with four different expressions.
Abstract: In the Fall of 2000, we collected a database of more than 40,000 facial images of 68 people. Using the Carnegie Mellon University 3D Room, we imaged each person across 13 different poses, under 43 different illumination conditions, and with four different expressions. We call this the CMU pose, illumination, and expression (PIE) database. We describe the imaging hardware, the collection procedure, the organization of the images, several possible uses, and how to obtain the database.
1,880 citations
TL;DR: The findings indicate that organizations are embedded in institutional networks and call for greater attention to be directed at understanding institutional pressures when investigating information technology innovations adoption.
Abstract: This study used institutional theory as a lens to understand the factors that enable the adoption of interorganizational systems. It posits that mimetic, coercive, and normative pressures existing in an institutionalized environment could influence organizational predisposition toward an information technology-based interorganizational linkage. Survey-based research was carried out to test this theory. Following questionnaire development, validation, and pretest with a pilot study, data were collected from the CEO, the CFO, and the CIO to measure the institutional pressures they faced and their intentions to adopt financial electronic data interchange (FEDI). A firm-level structural model was developed based on the CEO's, the CFO's, and the CIO's data. LISREL and PLS were used for testing the measurement and structural models respectively. Results showed that all three institutional pressures-mimetic pressures, coercive pressures, and normative pressures-had a significant influence on organizational intention to adopt FEDI. Except for perceived extent of adoption among suppliers, all other subconstructs were significant in the model. These results provide strong support for institutional-based variables as predictors of adoption intention for interorganizational linkages. These findings indicate that organizations are embedded in institutional networks and call for greater attention to be directed at understanding institutional pressures when investigating information technology innovations adoption.
1,601 citations
TL;DR: The behavior of the SVM classifier when these hyper parameters take very small or very large values is analyzed, which helps in understanding thehyperparameter space that leads to an efficient heuristic method of searching for hyperparameter values with small generalization errors.
Abstract: Support vector machines (SVMs) with the gaussian (RBF) kernel have been popular for practical use. Model selection in this class of SVMs involves two hyperparameters: the penalty parameter C and the kernel width σ. This letter analyzes the behavior of the SVM classifier when these hyperparameters take very small or very large values. Our results help in understanding the hyperparameter space that leads to an efficient heuristic method of searching for hyperparameter values with small generalization errors. The analysis also indicates that if complete model selection using the gaussian kernel has been conducted, there is no need to consider linear SVM.
1,586 citations
TL;DR: This review surveys select works of key groups in Asia developing chitin and chitosan materials for implantable biomedical applications and finds Japan has been the undisputed leader, but other Asian nations, namely Korea, Singapore, Taiwan and Thailand have also made notable contributions.
1,569 citations
TL;DR: In this article, the authors examined the impact of audit committee characteristics on corporate yields spreads as audit committees are the direct mechanism that boards use to monitor the financial accounting process and found that fully independent audit committee are associated with a significantly lower cost of debt financing.
Abstract: Creditor reliance on accounting-based debt covenants suggests that debtors are potentially concerned with board of director characteristics that influence the financial accounting process. In a sample of S&P 500 firms, we find that the cost of debt financing is inversely related to board independence and board size. We also examine the impact of audit committee characteristics on corporate yields spreads as audit committees are the direct mechanism that boards use to monitor the financial accounting process. We find that fully independent audit committees are associated with a significantly lower cost of debt financing. Similarly, yield spreads are also negatively related to audit committee size and the number of audit committee meetings. Overall, these results provide market-based evidence that boards and audit committees are important elements affecting the reliability of financial reports.
1,419 citations
TL;DR: A novel wet-chemical approach at 180 degrees C for the synthesis of monodispersed ZnO nanorods with high single-crystallinity with exceptionally high aspect ratio is reported, which should promise a future large-scale synthesis of ZNO nanostructures for many important applications in nanotechnology in a controlled manner.
Abstract: We report a novel wet-chemical approach at 180 °C for the synthesis of monodispersed ZnO nanorods with high single-crystallinity. The method has successfully brought the ZnO nanorod diameter from a reported 150 nm down to the 50 nm regime in this work. The aspect ratio of the synthesized nanorods achieved is exceptionally high (in the range of 30−40). This simple low-cost approach should promise us a future large-scale synthesis of ZnO nanostructures for many important applications in nanotechnology in a controlled manner.
1,330 citations
TL;DR: Computational experiments with linear optimization problems involving semidefinite, quadratic, and linear cone constraints (SQLPs) are discussed and computational results on problems from the SDPLIB and DIMACS Challenge collections are reported.
Abstract: This paper discusses computational experiments with linear optimization problems involving semidefinite, quadratic, and linear cone constraints (SQLPs). Many test problems of this type are solved using a new release of SDPT3, a Matlab implementation of infeasible primal-dual path-following algorithms. The software developed by the authors uses Mehrotra-type predictor-corrector variants of interior-point methods and two types of search directions: the HKM and NT directions. A discussion of implementation details is provided and computational results on problems from the SDPLIB and DIMACS Challenge collections are reported.
1,246 citations
TL;DR: SVM with adaptive parameters can both achieve higher generalization performance and use fewer support vectors than the standard SVM in financial forecasting.
Abstract: A novel type of learning machine called support vector machine (SVM) has been receiving increasing interest in areas ranging from its original application in pattern recognition to other applications such as regression estimation due to its remarkable generalization performance. This paper deals with the application of SVM in financial time series forecasting. The feasibility of applying SVM in financial forecasting is first examined by comparing it with the multilayer back-propagation (BP) neural network and the regularized radial basis function (RBF) neural network. The variability in performance of SVM with respect to the free parameters is investigated experimentally. Adaptive parameters are then proposed by incorporating the nonstationarity of financial time series into SVM. Five real futures contracts collated from the Chicago Mercantile Market are used as the data sets. The simulation shows that among the three methods, SVM outperforms the BP neural network in financial forecasting, and there are comparable generalization performance between SVM and the regularized RBF neural network. Furthermore, the free parameters of SVM have a great effect on the generalization performance. SVM with adaptive parameters can both achieve higher generalization performance and use fewer support vectors than the standard SVM in financial forecasting.
916 citations
TL;DR: Revised formulas for summarizing retinal vessel diameters measured from fundus photographs are described, which offer the advantages of being more robust against variability in the number of vessels observed, being independent of image scale, and being easier to implement.
Abstract: Background/Purpose. Recent findings suggest that an objective assessment of retinal vessel caliber from fundus photographs provide information about the association of microvascular characteristics with macrovascular disease. Current methods used to quantify retinal vessel caliber, introduced by Parr(1,2) and Hubbard,(3) are not independent of scale and are affected by the number of vessels. To improve upon these methods we introduce revised formulas for quantifying vessel caliber. Methods. Revised formulas were estimated using retinal vessel measurements from 44 young adults free of hypertension and diabetes. Comparisons between the two methods were done using digitized photographs from 4926 participants at the baseline examination of the Beaver Dam Eye Study (BDES), an ongoing population-based cohort study initiated in 1987. Individual arterioles and venules were measured using semi-automated computer software from which summary measures were calculated. Results. Correlation coefficients between the Parr-Hubbard and revised formulas were high (Pearson correlation coefficients ranging from 0.94 to 0.98). Both arteriolar and venular caliber significantly increased with an increasing number of vessels measured using the Parr-Hubbard formulas (p 0.50). Conclusions. We describe revised formulas for summarizing retinal vessel diameters measured from fundus photographs to be used in future studies and analyses. The revised formulas correlate highly with the previously used Parr-Hubbard formulas, but offer the advantages of being more robust against variability in the number of vessels observed, being independent of image scale, and being easier to implement.
TL;DR: In this article, the authors test overreaction theories of short-run momentum and long-run reversal in the cross section of stock returns and find that macroeconomic factors are unable to explain momentum profits after simple methodological adjustments to take account of microstructure concerns.
Abstract: We test overreaction theories of short-run momentum and long-run reversal in the cross section of stock returns. Momentum profits depend on the state of the market, as predicted. From 1929 to 1995, the mean monthly momentum profit following positive market returns is 0.93 percent, whereas the mean profit following negative market returns is negative 0.37 percent. The up-market momentum reverses in the long-run. Our results are robust to the conditioning information in macroeconomic factors. Moreover, we find that macroeconomic factors are unable to explain momentum profits after simple methodological adjustments to take account of microstructure concerns.
TL;DR: Vitamin E TPGS has great advantages for the manufacture of polymeric nanoparticles for controlled release of paclitaxel and other anti-cancer drugs and could be a novel surfactant as a matrix material when blended with other biodegradable polymers.
TL;DR: This paper proposes a text independent method of emotion classification of speech that makes use of short time log frequency power coefficients (LFPC) to represent the speech signals and a discrete hidden Markov model (HMM) as the classifier.
19 Nov 2003
TL;DR: A more principled approach to solving the problem of building text classifiers using positive and unlabeled examples based on a biased formulation of SVM is proposed, and it is shown experimentally that it is more accurate than the existing techniques.
Abstract: We study the problem of building text classifiers using positive and unlabeled examples. The key feature of this problem is that there is no negative example for learning. Recently, a few techniques for solving this problem were proposed in the literature. These techniques are based on the same idea, which builds a classifier in two steps. Each existing technique uses a different method for each step. We first introduce some new methods for the two steps, and perform a comprehensive evaluation of all possible combinations of methods of the two steps. We then propose a more principled approach to solving the problem based on a biased formulation of SVM, and show experimentally that it is more accurate than the existing techniques.
TL;DR: The mechanical deformation characteristics of living cells are known to influence strongly their chemical and biological functions and the onset, progression and consequences of a number of human diseases and potential applications of the optical tweezers method are highlighted.
Abstract: The mechanical deformation characteristics of living cells are known to influence strongly their chemical and biological functions and the onset, progression and consequences of a number of human diseases. The mechanics of the human red blood cell (erythrocyte) subjected to large deformation by optical tweezers forms the subject of this paper. Video photography of the cell deformed in a phosphate buffered saline solution at room temperature during the imposition of controlled stretching forces, in the tens to several hundreds picoNewton range, is used to assess experimentally the deformation characteristics. The mechanical responses of the cell during loading and upon release of the optical force are then analysed to extract the elastic properties of the cell membrane by recourse to several different constitutive formulations of the elastic and viscoelastic behavior within the framework of a fully three-dimensional finite element analysis. A parametric study of various geometric, loading and structural factors is also undertaken in order to develop quantitative models for the mechanics of deformation by means of optical tweezers. The outcome of the experimental and computational analyses is then compared with the information available on the mechanical response of the red blood cell from other independent experimental techniques. Potential applications of the optical tweezers method described in this paper to the study of mechanical deformation of living cells under different stress states and in response to the progression of some diseases are also highlighted.
TL;DR: Extrapolations of the observed and inferred local extinction data imply that the current unprecedented rate of habitat destruction in Southeast Asia will result in the loss of 13–42% of regional populations over the next century, at least half of which will represent global species extinctions.
Abstract: The looming mass extinction of biodiversity in the humid tropics is a major concern for the future, yet most reports of extinctions in these regions are anecdotal or conjectural, with a scarcity of robust, broad-based empirical data. Here we report on local extinctions among a wide range of terrestrial and freshwater taxa from Singapore (540 km2) in relation to habitat loss exceeding 95% over 183 years. Substantial rates of documented and inferred extinctions were found, especially for forest specialists, with the greatest proportion of extinct taxa (34-87%) in butterflies, fish, birds and mammals. Observed extinctions were generally fewer, but inferred losses often higher, in vascular plants, phasmids, decapods, amphibians and reptiles (5-80%). Forest reserves comprising only 0.25% of Singapore's area now harbour over 50% of the residual native biodiversity. Extrapolations of the observed and inferred local extinction data, using a calibrated species-area model, imply that the current unprecedented rate of habitat destruction in Southeast Asia will result in the loss of 13-42% of regional populations over the next century, at least half of which will represent global species extinctions.
TL;DR: The process by which the initial structure with random spatial composition fluctuations turns into an alloy (solid solution) with homogeneous composition is clearly demonstrated by the temporal evolution of the PL spectra during the annealing progress, which confirms the formation of homogeneous nanoalloys.
Abstract: High-quality alloyed ZnxCd1-xS nanocrystals have been synthesized at high temperature by the reaction of a mixture of CdO- and ZnO-oleic acid complexes with sulfur in the noncoordinating solvent octadecene system. A series of monodisperse wurtzite ZnxCd1-xS (x = 0.10, 0.25, 0.36, 0.53) nanocrystals were obtained with corresponding particle radii of 4.0, 3.2, 2.9, and 2.4 nm, respectively. With the increase of the Zn content, their photoluminescence (PL) spectra blue-shift systematically across the visible spectrum from 474 to 391 nm, indicating the formation of the alloyed nanocrystals. The alloy structure is also supported by the characteristic X-ray diffraction (XRD) patterns of these nanoalloys with different Zn mole fractions, in which their diffraction peaks systematically shift to larger angles as the Zn content increases. The lattice parameter c measured from XRD patterns decreases linearly with the increase of Zn content. This trend is consistent with Vegard's law, which further confirms the forma...
28 Jul 2003
TL;DR: This paper proposes to use a special kernel function called the tree kernel to enable the SVM to take advantage of the syntactic structures of questions, and describes how the tree Kernel can be computed efficiently by dynamic programming.
Abstract: Question classification is very important for question answering. This paper presents our research work on automatic question classification through machine learning approaches. We have experimented with five machine learning algorithms: Nearest Neighbors (NN), Naive Bayes (NB), Decision Tree (DT), Sparse Network of Winnows (SNoW), and Support Vector Machines (SVM) using two kinds of features: bag-of-words and bag-of-ngrams. The experiment results show that with only surface text features the SVM outperforms the other four methods for this task. Further, we propose to use a special kernel function called the tree kernel to enable the SVM to take advantage of the syntactic structures of questions. We describe how the tree kernel can be computed efficiently by dynamic programming. The performance of our approach is promising, when tested on the questions from the TREC QA track.
TL;DR: The empirically study the usefulness of several simple performance measures that are inexpensive to compute (in the sense that they do not require expensive matrix operations involving the kernel matrix) for tuning SVM hyperparameters.
TL;DR: In this paper, a microencapsulated phase change material (PCM) for solar thermal-energy storage capacities has been investigated in terms of encapsulation efficiency and energy storage and release capacity.
TL;DR: A review of surface modification techniques for polymers with graft chains can be found in this paper, focusing on grafting methods as well as the structure and function of grafted surfaces.
TL;DR: In this paper, the authors find support for the role of experiential learning in the international expansion process by extending the stages model of internationalization to incorporate a sophisticated consideration of temporal and cross-national variation in the credibility of the policy environment.
Abstract: We find support for the role of experiential learning in the international expansion process by extending the stages model of internationalization to incorporate a sophisticated consideration of temporal and cross-national variation in the credibility of the policy environment. Using a sample of 3857 international expansions of 665 Japanese manufacturing firms, we build on the concepts of uncertainty and experiential learning, to show that firms that had gathered relevant types of international experience were less sensitive to the deterring effect of uncertain policy environments on investment. One implication of our results is that research on international strategy should emphasize understanding the political institutions that constrain or enable political actors, just as entry mode research has done. A second implication is that research in the stages model of internationalization should give the same weight to the policy environment as a source of uncertainty to a firm, as it has given to cultural, social and market institutions. Copyright © 2003 John Wiley & Sons, Ltd.
TL;DR: Identification of the pharmacological sites has potential for exploitation in development of novel systems useful for 'delivering' specific proteins to a particular target tissue or organ and research in this field will provide a lot of exciting opportunities.
TL;DR: The mechanism of the high luminescence efficiency and stability of Zn(x)Cd(1-x)Se nanocrystals is explored and a composition-tunable emission across most of the visible spectrum has been demonstrated by a systematic blue-shift in emission wavelength.
Abstract: High-quality ZnxCd1-xSe nanocrystals have been successfully prepared at high temperature by incorporating stoichiometric amounts of Zn and Se into pre-prepared CdSe nanocrystals. With increasing Zn content, a composition-tunable emission across most of the visible spectrum has been demonstrated by a systematic blue-shift in emission wavelength. The photoluminescence (PL) properties for the obtained ZnxCd1-xSe nanocrystals (PL efficiency of 70−85%, fwhm = 22−30 nm) are comparable to those for the best reported CdSe-based QDs. In particular, they also have good PL properties in the blue spectral range. Moreover, the alloy nanocrystals can retain their high luminescence (PL efficiency of over 40%) when dispersed in aqueous solutions and maintain a symmetric peak shape and spectral position under rigorous experimental conditions. A rapid alloying process was observed at a temperature higher than “alloying point”. The mechanism of the high luminescence efficiency and stability of ZnxCd1-xSe nanocrystals is exp...
TL;DR: The results show that attitudinal and social factors rather than perceived behavioral control factors play a significant role in influencing intentions to adopt a WAP-enabled mobile phone.
Abstract: This paper examines the attitudinal, social and perceived behavior control factors that are associated with the adoption of WAP-enabled mobile phones among Internet users. An online questionnaire is used to gather data. The results show that attitudinal and social factors rather than perceived behavioral control factors play a significant role in influencing intentions to adopt a WAP-enabled mobile phone. In particular, perceptions of relative advantage, risk, and image are found to influence adoption intentions. In addition, reference groups too play an important role in shaping adoption intentions. Implications of results and directions for future research are examined.
TL;DR: SVMProt shows a certain degree of capability for the classification of distantly related proteins and homologous proteins of different function and thus may be used as a protein function prediction tool that complements sequence alignment methods.
Abstract: Prediction of protein function is of significance in studying biological processes. One approach for function prediction is to classify a protein into functional family. Support vector machine (SVM) is a useful method for such classification, which may involve proteins with diverse sequence distribution. We have developed a web-based software, SVMProt, for SVM classification of a protein into functional family from its primary sequence. SVMProt classification system is trained from representative proteins of a number of functional families and seed proteins of Pfam curated protein families. It currently covers 54 functional families and additional families will be added in the near future. The computed accuracy for protein family classification is found to be in the range of 69.1–99.6%. SVMProt shows a certain degree of capability for the classification of distantly related proteins and homologous proteins of different function and thus may be used as a protein function prediction tool that complements sequence alignment methods. SVMProt can be accessed at http://jing.cz3.nus.edu.sg/cgi-bin/svmprot.cgi.
TL;DR: In this article, a field measurement was conducted in Singapore to investigate the thermal impacts of rooftop garden, and it has been confirmed that rooftop gardens contribute thermal benefits to both buildings and their surrounding environments.
TL;DR: In this article, the numerical solution of the time-dependent Gross-Pitaevskii equation (GPE) describing a Bose-Einstein condensate (BEC) at zero or very low temperature is studied.
TL;DR: The experiment shows that SVM by feature extraction using PCA, KPCA or ICA can perform better than that without feature extraction, and among the three methods, there is the best performance in K PCA feature extraction; followed by ICA feature extraction.