Showing papers by "National Taiwan University of Science and Technology published in 2007"

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.

Transitional Markov Chain Monte Carlo Method for Bayesian Model Updating, Model Class Selection, and Model Averaging

Abstract: This paper presents a newly developed simulation-based approach for Bayesian model updating, model class selection, and model averaging called the transitional Markov chain Monte Carlo (TMCMC) approach. The idea behind TMCMC is to avoid the problem of sampling from difficult target probability density functions (PDFs) but sampling from a series of intermediate PDFs that converge to the target PDF and are easier to sample. The TMCMC approach is motivated by the adaptive Metropolis–Hastings method developed by Beck and Au in 2002 and is based on Markov chain Monte Carlo. It is shown that TMCMC is able to draw samples from some difficult PDFs (e.g., multimodal PDFs, very peaked PDFs, and PDFs with flat manifold). The TMCMC approach can also estimate evidence of the chosen probabilistic model class conditioning on the measured data, a key component for Bayesian model class selection and model averaging. Three examples are used to demonstrate the effectiveness of the TMCMC approach in Bayesian model updating, ...

Adoption of the mobile Internet: An empirical study of multimedia message service (MMS)

Abstract: Multimedia message service (MMS) provides more multimedia communication with entertainment effects than current text-based short message service (SMS). While many reports indicate that the mobile Internet market will be huge, little is known about whether people will accept MMS. This study applies innovation diffusion theory to examine the factors that influence the adoption of MMS. The proposed model was empirically evaluated by using survey data collected from 207 users concerning their perceptions of MMS. The findings indicate that perceptions of use were different over innovation diffusion stages. Specifically, there was a significant difference between potential adopters and users. The results may provide further insights into MMS marketing strategies.

Interorganizational Governance Value Creation: Coordinating for Information Visibility and Flexibility in Supply Chains*

Abstract: More thoroughly understanding how interorganizational governance value can be created by information technology and other governance mechanisms is critical for supply chain management. Based primarily on transaction-cost economics and supplemented by the resource-based view, this study investigates how interorganizational governance (i.e., relational governance and virtual integration) can create value (i.e., information visibility and supply chain flexibility) in the supply chain context. The findings show that both relational governance and virtual integration benefit information visibility. Those results also support both direct and indirect (via information visibility) effects of relational governance on supply chain flexibility. Although failing to affect supply chain flexibility directly, virtual integration can still improve supply chain flexibility with its ability to enhance information visibility. Thus, interorganizational governance mechanisms emphasizing both control and collaboration can influence the gain from collaborationspecific capabilities, leading to the competitive advantage of a supply chain. The results of the study suggest that firms can gain greater supply chain flexibility within existing interfirm relationships by enhancing information visibility through virtual integration and relational governance.

Fuzzy risk analysis based on the ranking of generalized trapezoidal fuzzy numbers

Abstract: In this paper, we present a new method for fuzzy risk analysis based on the ranking of generalized trapezoidal fuzzy numbers. The proposed method considers the centroid points and the standard deviations of generalized trapezoidal fuzzy numbers for ranking generalized trapezoidal fuzzy numbers. We also use an example to compare the ranking results of the proposed method with the existing centroid-index ranking methods. The proposed ranking method can overcome the drawbacks of the existing centroid-index ranking methods. Based on the proposed ranking method, we also present an algorithm to deal with fuzzy risk analysis problems. The proposed fuzzy risk analysis algorithm can overcome the drawbacks of the one we presented in [7].

Reduced Support Vector Machines: A Statistical Theory

Abstract: In dealing with large data sets, the reduced support vector machine (RSVM) was proposed for the practical objective to overcome some computational difficulties as well as to reduce the model complexity. In this paper, we study the RSVM from the viewpoint of sampling design, its robustness, and the spectral analysis of the reduced kernel. We consider the nonlinear separating surface as a mixture of kernels. Instead of a full model, the RSVM uses a reduced mixture with kernels sampled from certain candidate set. Our main results center on two major themes. One is the robustness of the random subset mixture model. The other is the spectral analysis of the reduced kernel. The robustness is judged by a few criteria as follows: 1) model variation measure; 2) model bias (deviation) between the reduced model and the full model; and 3) test power in distinguishing the reduced model from the full one. For the spectral analysis, we compare the eigenstructures of the full kernel matrix and the approximation kernel matrix. The approximation kernels are generated by uniform random subsets. The small discrepancies between them indicate that the approximation kernels can retain most of the relevant information for learning tasks in the full kernel. We focus on some statistical theory of the reduced set method mainly in the context of the RSVM. The use of a uniform random subset is not limited to the RSVM. This approach can act as a supplemental algorithm on top of a basic optimization algorithm, wherein the actual optimization takes place on the subset-approximated data. The statistical properties discussed in this paper are still valid

Factors influencing online music purchase intention in Taiwan

Abstract: Purpose – The purpose of this paper is to provide an explanation of factors influencing online music purchase intention of Taiwanese early adopter of online music, which can help the online music practitioners of Taiwan to develop better market strategies.Design/methodology/approach – An empirical survey was used to test the hypotheses. Data were collected from a total of 302 online Taiwanese early adopters of online music. A structural equation modeling (SEM) is proposed to assess the relationships of the research model.Finding – The findings in this paper show that the perceived value of online music is a significant factor in predicting the purchaser intention of buying online music in Taiwan. Also, the beneficial factor of the perceived usefulness and playfulness are identified in addition to the sacrificing factor of the perceived price for assessing the value. Moreover, purchasers and potential purchasers differ in the determinants underlying the perceptions of value, which customers hold towards on...

Material properties of portland cement paste with nano-montmorillonite

Abstract: The nano-montmorillonite, which has characteristics of high aspect ratio and interaction between polymer chains and dispersed nanolayers, has been widely used in the development of new reinforced nanocomposite polymers to improve their mechanical properties. Since a potential pozzolanic reaction may occur between Portland cement paste and high amount of silicon dioxide (SiO2) in nano-montmorillonite, the effects of introduction of montmorillonite to Portland cement-based material on the improvement of matrix properties of cement paste is of great interest in the construction industry. In this study, a liquid-form of nano-montmorillonite particle with a planar diameter of about 100 nm were incorporated into the Portland cement paste at five different dosages and analyzed at four different ages to identify the nanosizing effects on material properties of such cement-based composite. Experimental results show that the composite with 0.60% and 0.40% of added nano-montmorillonite by weight of cement have the optimum compressive strength and permeability coefficient, respectively, in which the increase of compressive strength is about 13.24%, and the decrease of permeability coefficient about 49.95%. Microstructural properties through the analyses of XRD, DSC, NMR, and MIP also indicate that the microstructures of cement paste with nano-montmorillonite contain more dense solid material and more stable bonding framework.

LED backlight driving system for large-scale of LCD panels

Abstract: This paper presents a LED backlight driving system for large-scale of LCD panels. High efficiency, high power factor, circuit simplicity, and low cost can be achieved by using a single-stage charge-pump asymmetrical-half-bridge converter. To regulate the LED current and brightness for the LED backlight system, some current sharing methods are presented and compared. The requirements for the current sharing and luminance balance among paralleled LED arrays can be satisfied while current ripple is eliminated significantly. Because of the addition of bypass diodes, an alternative current path can be offered when a single LED fails. The LED array will not distinguish. A laboratory prototype has been built and tested. The simulation and experimental results are shown to verify the feasibility of the proposed method.

LED Backlight Driving System for Large-Scale LCD Panels

Abstract: This paper presents an LED backlight driving system for large-scale liquid crystal display panels. High efficiency, high power factor, circuit simplicity, and low cost can be achieved by using a single-stage charge-pump asymmetrical half-bridge converter. To regulate the LED current and brightness for the LED backlight system, some current sharing methods are presented and compared. The requirements for the current sharing and luminance balance among paralleled LED arrays can be satisfied while current ripple is eliminated significantly. Because of the addition of bypass diodes, an alternative current path can be offered when a single LED fails. The LED array will not be extinguished. Reliability of the LED backlight system can thus be improved effectively. A laboratory prototype has been built and tested. The simulation and experimental results are shown to verify the feasibility of the proposed method.

A New Planar Artificial Transmission Line and Its Applications to a Miniaturized Butler Matrix

Abstract: A miniaturized quadrature hybrid coupler, a rat-race coupler, and a 4 times 4 Butler matrix based on a newly proposed planar artificial transmission line are presented in this paper for application in ultra-high-frequency (UHF) radio-frequency identification (RFID) systems. This planar artificial transmission line is composed of microstrip quasi-lumped elements and their discontinuities and is capable of synthesizing microstrip lines with various characteristic impedances and electrical lengths. At the center frequency of the UHF RFID system, the occupied sizes of the proposed quadrature hybrid and rat-race couplers are merely 27% and 9% of those of the conventional designs. The miniaturized couplers demonstrate well-behaved wideband responses with no spurious harmonics up to two octaves. The measured results reveal excellent agreement with the simulations. Additionally, a 4 times 4 Butler matrix, which may occupy a large amount of circuit area in conventional designs, has been successfully miniaturized with the help of the proposed artificial transmission line. The circuit size of the Butler matrix is merely 21% of that of a conventional design. The experimental results show that the proposed Butler matrix features good phase control, nearly equal power splitting, and compact size and is therefore applicable to the reader modules in various RFID systems.

Understanding intention to continuously share information on weblogs

Abstract: Purpose – Weblogs, or blogs, have been a significant new development in recent years. Many businesses have begun using blogs to stimulate discussions, garner ideas, and provide further visibility. Thus, blog usage as well as individual motives behind continued usage is an important area of research. This research aims to examine behavioral motivations underlying individual intention to keep using blogs.Design/methodology/approach – Based on social cognitive theory and related technology adoption literature, this study considered knowledge self‐efficacy, subjective norms, feedback, and personal outcome expectations as the determinants of continuing to share information on blogs. The proposed model was empirically evaluated using randomized survey data collected from 155 users of a popular web site in Taiwan (WRETCH blog).Findings – The results revealed that although both self‐efficacy and personal outcome expectations affected directly the intention of sharing information on blogs, personal outcome expecta...

High School Students’ Informal Reasoning on a Socio‐scientific Issue: Qualitative and quantitative analyses

Abstract: Recently, the significance of learners’ informal reasoning on socio‐scientific issues has received increasing attention among science educators. To gain deeper insights into this important issue, an integrated analytic framework was developed in this study. With this framework, 71 Grade 10 students’ informal reasoning about nuclear energy usage was explored qualitatively and quantitatively. It was found that the students in this study tended to process reasoning from multiple perspectives, and most of them were prone to make evidence‐based decisions. However, less than 40% of the participants were able to construct rebuttals against counter‐arguments. It was also revealed that students’ abundant usage of supportive arguments did not guarantee for their counter‐argument construction as well as rebuttal construction, but their usage of counter‐arguments might act as precursors to their construction of rebuttals. In addition, learners’ usage of multiple reasoning modes might help them propose more arguments ...

An Assessment of Testing-Effort Dependent Software Reliability Growth Models

Abstract: Over the last several decades, many Software Reliability Growth Models (SRGM) have been developed to greatly facilitate engineers and managers in tracking and measuring the growth of reliability as software is being improved. However, some research work indicates that the delayed S-shaped model may not fit the software failure data well when the testing-effort spent on fault detection is not a constant. Thus, in this paper, we first review the logistic testing-effort function that can be used to describe the amount of testing-effort spent on software testing. We describe how to incorporate the logistic testing-effort function into both exponential-type, and S-shaped software reliability models. The proposed models are also discussed under both ideal, and imperfect debugging conditions. Results from applying the proposed models to two real data sets are discussed, and compared with other traditional SRGM to show that the proposed models can give better predictions, and that the logistic testing-effort function is suitable for incorporating directly into both exponential-type, and S-shaped software reliability models

Ultrawideband Band-Notched Folded Strip Monopole Antenna

Abstract: We propose a new band-notched folded strip monopole antenna for ultrawideband applications. This antenna is composed of a forked-shape radiator and a 50 Omega microstrip line. To achieve band-rejected filtering property at the WLAN bands, the forked-shape strips are folded back and result in a pair of coupled lines on the radiator. The length and gap width of the coupled lines primarily determine the notched frequency of the antenna. Based on the band-notched resonance, an equivalent circuit model is proposed for the antenna and the calculated antenna input admittance agrees with the full-wave simulation data. With the help of the dimensionless normalized antenna transfer function, the radiation characteristics are investigated thoroughly. The transmission responses of a transceiving antenna system and their corresponding transient analysis are discussed at the end of this paper.

Optimization of Plasma Arc Welding Parameters by Using the Taguchi Method with the Grey Relational Analysis

Abstract: The optimal parameters process of plasma arc welding (PAW) by the Taguchi method with Grey relational analysis is studied. The Grey relational grade is used to find optimal PAW parameters with multiple response performance characteristics. The welding parameters (welding current, welding speed, plasma gas flow rate, and torch stand-off) are optimized with consideration of the multiple response performance characteristics (the penetration of root, the weld groove width, and the weld pool undercut). As a result, the improvement percentage of the Grey relational grade with the multiple performance characteristics is 31.8%. It is shown that the multiple response performance characteristics are greatly improved through this study.

Temperature prediction and TAIFEX forecasting based on fuzzy logical relationships and genetic algorithms

Abstract: In this paper, we present a new method for temperature prediction and the TAIFEX forecasting based on fuzzy logical relationships and genetic algorithms. The proposed method constructs two-factors high-order fuzzy logical relationships based on the historical data and uses genetic algorithms to adjust the length of each interval in the universe of discourse for temperature prediction and the TAIFEX forecasting to increase the forecasting accuracy rate. The proposed method gets a higher forecasting accuracy rate than the existing methods.

Teachers' scientific epistemological views: The coherence with instruction and students' views***

Abstract: Research about the relationship between teachers' scientific epistemological views (SEVs) and science instruction is often an important issue for many science educators. This study, by collecting research data from four Taiwanese science teachers, their students, and classroom observations, was carried out to examine the coherences between teachers' SEVs and their (1) teaching beliefs, (2) instructional practices, (3) students' SEVs, and (4) students' perceptions toward actual science learning environments. The findings suggested adequate coherences between teachers' SEVs and their teaching beliefs as well as instructional practices. The teachers with relatively positivist-aligned SEVs tended to draw attention to students' science scores in tests and allocate more instructional time on teacher-directed lectures, tutorial problem practices, or in-class examinations, implying a more passive or rote perspective about learning science. In contrast, teachers with constructivist-oriented SEVs tended to focus on student understanding and application of scientific concepts and they adopted more time on student inquiry activities or interactive discussion. These findings are quite consistent with the results about the coherence between teachers' SEVs and students' perceptions toward science learning environments, suggesting that the constructivist-oriented SEVs appeared to foster the creation of more constructivist-oriented science learning environments. Finally, although this study provided some evidence that teachers' SEVs were likely related to their students' SEVs, the teachers' SEVs and those of their students were not obviously coherent. © 2006 Wiley Periodicals, Inc. Sci Ed91:222–243, 2007

A Merged Fuzzy Neural Network and Its Applications in Battery State-of-Charge Estimation

Abstract: To solve learning problems with vast number of inputs, this paper proposes a novel learning structure merging a number of small fuzzy neural networks (FNNs) into a hierarchical learning structure called a merged-FNN. In this paper, the merged-FNN is proved to be a universal approximator. This computing approach uses a fusion of FNNs using B-spline membership functions (BMFs) with a reduced-form genetic algorithm (RGA). RGA is employed to tune all free parameters of the merged-FNN, including both the control points of the BMFs and the weights of the small FNNs. The merged-FNN can approximate a continuous nonlinear function to any desired degree of accuracy. For a practical application, a battery state-of-charge (BSOC) estimator, which is a twelve input, one output system, in a lithium-ion battery string is proposed to verify the effectiveness of the merged-FNN. From experimental results, the learning ability of the newly proposed merged-FNN with RGA is superior to that of the traditional neural networks with back-propagation learning.

Remote-Controllable Power Outlet System for Home Power Management

Abstract: In this paper we describe the Wireless Power-Controlled Outlet Module (WPCOM) with a scalable mechanism for home power management which we have developed. The WPCOM integrates the multiple AC power sockets and a simple low-power microcontroller into a power outlet to switch the power of the sockets ON/OFF and to measure the power consumption of plugged electric home appliances. Our WPCOM consists of six scalable modules, that is, the Essential Control Module, the Bluetooth Module, the GSM Module, the Ethernet Module, the SD Card Module and the Power Measuring Module, which together provide an indoor wireless, and an outdoor remote control and monitor of electric home appliances. We have designed a PDA control software and remote control software which support the Graphic User Interface, thus allowing the user to easily monitor the electric home appliances through the PDA and the Internet individually. In addition, we use a Short Message Service to achieve control and monitoring through a GSM cellular mobile phone for remote use anytime and anywhere.