Showing papers by "National Taiwan University of Science and Technology published in 2008"

Particle swarm optimization for parameter determination and feature selection of support vector machines

Abstract: Support vector machine (SVM) is a popular pattern classification method with many diverse applications. Kernel parameter setting in the SVM training procedure, along with the feature selection, significantly influences the classification accuracy. This study simultaneously determines the parameter values while discovering a subset of features, without reducing SVM classification accuracy. A particle swarm optimization (PSO) based approach for parameter determination and feature selection of the SVM, termed PSO+SVM, is developed. Several public datasets are employed to calculate the classification accuracy rate in order to evaluate the developed PSO+SVM approach. The developed approach was compared with grid search, which is a conventional method of searching parameter values, and other approaches. Experimental results demonstrate that the classification accuracy rates of the developed approach surpass those of grid search and many other approaches, and that the developed PSO+SVM approach has a similar result to GA+SVM. Therefore, the PSO+SVM approach is valuable for parameter determination and feature selection in an SVM.

Criteria, Strategies and Research Issues of Context-Aware Ubiquitous Learning

Abstract: Recent progress in wireless and sensor technologies has lead to a new development of learning environments, called context-aware ubiquitous learning environment, which is able to sense the situation of learners and provide adaptive supports. Many researchers have been investigating the development of such new learning environments; nevertheless, the criteria of establishing a context-aware ubiquitous learning environment have not yet been clearly defined, not to mention the strategies of conducting effective learning activities. To resolve these problems, this paper presents the basic criteria, strategies, and research issues of context-aware ubiquitous learning, and identifies the necessary check items as well for the development of such learning environment. Illustrative examples of conducting context-aware ubiquitous learning activities and the requirements of setting up such learning environment are also presented at the end of this paper.

Parameter determination of support vector machine and feature selection using simulated annealing approach

Abstract: Support vector machine (SVM) is a novel pattern classification method that is valuable in many applications. Kernel parameter setting in the SVM training process, along with the feature selection, significantly affects classification accuracy. The objective of this study is to obtain the better parameter values while also finding a subset of features that does not degrade the SVM classification accuracy. This study develops a simulated annealing (SA) approach for parameter determination and feature selection in the SVM, termed SA-SVM. To measure the proposed SA-SVM approach, several datasets in UCI machine learning repository are adopted to calculate the classification accuracy rate. The proposed approach was compared with grid search which is a conventional method of performing parameter setting, and various other methods. Experimental results indicate that the classification accuracy rates of the proposed approach exceed those of grid search and other approaches. The SA-SVM is thus useful for parameter determination and feature selection in the SVM.

A decision support system for supplier selection based on a strategy-aligned fuzzy SMART approach

Abstract: This study presents a strategy-aligned fuzzy simple multiattribute rating technique (SMART) approach for solving the supplier/vendor selection problem from the perspective of strategic management of the supply chain (SC). The majority of supplier rating systems obtained their optimal solutions without considering firm operations management (OM)/SC strategy. The proposed system utilizes OM/SC strategy to identify supplier selection criteria. A fuzzy SMART is applied to evaluate the alternative suppliers, and deals with the ratings of both qualitative and quantitative criteria. The final decision-maker incorporates the supply risks of individual suppliers into final decision making. Finally, an empirical study is conducted to demonstrate the procedure of the proposed system and identify the suitable supplier(s).

Development of an adaptive learning system with two sources of personalization information

Abstract: Previous research of adaptive learning mainly focused on improving student learning achievements based only on single-source of personalization information, such as learning style, cognitive style or learning achievement. In this paper, an innovative adaptive learning approach is proposed by basing upon two main sources of personalization information, that is, learning behavior and personal learning style. To determine the initial learning styles of the students, the [Keefe, J. W. (1987). Learning Styles: Theory and Practice. Reston, VA: National Association of Secondary School Principals.] questionnaire is employed in our approach. To more precisely reflect the learning behaviors of each student, the interactions and learning results of each student are analyzed when adjusting the subject materials. Based on the innovative approach, an adaptive learning system has been developed; moreover, an experiment was conducted to evaluate the performance of our approach. By analyzing the results from three groups of students using different adaptive learning approaches, it can be found that the innovative approach is helpful in improving both the learning achievement and learning efficiency of individual students.

The role of feedback and self-efficacy on web-based learning: The social cognitive perspective

Abstract: The social cognitive perspective of self-regulated learning suggests that effective learning is determined by the interactions among personal, behavioral, and environmental influences; particularly, high self-regulated learners hold higher motivation (personal), apply better learning strategies (behavioral) and respond to environmental demand more appropriately (environmental). The study thus uses the social cognitive perspective to explore the role of self-efficacy (personal), student feedback behavior, use of learning strategies (behavioral), performance and receiving feedback (environmental) in Web-based learning. There were 76 university students participated in this study. Both quantitative and qualitative methods were applied for data analysis. The results supported that self-efficacy predicted student use of learning strategies and related to elaborated feedback behavior (personal->behavioral). High self-efficacy students applied more high-level learning strategies, such as elaborative strategy and critical thinking. Students who provided elaborated feedback also had higher self-efficacy than those who did not. Moreover, receiving elaborative feedback significantly promoted student self-efficacy (environmental->personal), while receiving knowledge of correct response improved student performance. However, the results indicated that feedback behaviors did not predict academic performance, which may be interfered by modeling effects.

The role of experience and innovation characteristics in the adoption and continued use of e-learning websites

Abstract: With the advent of e-learning technologies in the past decade, the accessibility of training, teaching, and learning has drastically increased. The challenge for the education enterprise now is how to attract learners to their e-learning services. In this study, a technology adoption model is developed to predict the users' intention of adoption and their continued use behavior. The results show significant evidence in support of the hypothesis. The findings indicate that perceptions of relative advantage and compatibility are significantly related to users' intention to use e-learning. Also, the intention is significantly related to their actual use of e-learning. Furthermore, the technology adoption of learners with prior e-learning experience is different from those without prior e-learning experience. These findings may contribute to deeper understanding of e-learning users' perceptions in terms of adoption and continued use behavior.

Highly sensitive ZnO nanowire CO sensors with the adsorption of Au nanoparticles

Abstract: In this study, the growth of high density single-crystalline ZnO nanowires on patterned ZnO:Ga/SiO2/Si templates was reported. We also adsorbed Au nanoparticles onto nanowire surfaces and fabricated ZnO nanowire CO sensors. With 50 ppm CO gas, it was found that we could enhance the device sensitivities at 350 ◦ C from 4.2% to 46.5% by the adsorption of Au nanoparticles. It was also found that measured sensitivities were around 30%, 37%, 46.5% and 53% when concentration of the injected CO gas was 5, 20, 50 and 100 ppm, respectively. (Some figures in this article are in colour only in the electronic version)

An Efficient Watermarking Method Based on Significant Difference of Wavelet Coefficient Quantization

Abstract: This paper proposes a blind watermarking algorithm based on the significant difference of wavelet coefficient quantization for copyright protection. Every seven nonoverlap wavelet coefficients of the host image are grouped into a block. The largest two coefficients in a block are called significant coefficients in this paper and their difference is called significant difference. We quantized the local maximum wavelet coefficient in a block by comparing the significant difference value in a block with the average significant difference value in all blocks. The maximum wavelet coefficients are so quantized that their significant difference between watermark bit 0 and watermark bit 1 exhibits a large energy difference which can be used for watermark extraction. During the extraction, an adaptive threshold value is designed to extract the watermark from the watermarked image under different attacks. We compare the adaptive threshold value to the significant difference which was quantized in a block to determine the watermark bit. The experimental results show that the proposed method is quite effective against JPEG compression, low-pass filtering, and Gaussian noise; the PSNR value of a watermarked image is greater than 40 dB.

A Low-Cost 60-GHz Switched-Beam Patch Antenna Array With Butler Matrix Network

Abstract: In this letter, a low-cost 60-GHz switched-beam patch antenna array with Butler matrix network is developed and experimentally demonstrated. In order to improve integration with the patch elements, a 4 times 4 planar Butler matrix is implemented in a low-dielectric substrate. The four rectangular patches fed by inset microstrip lines are connected to the outputs of the Butler matrix for the 60-GHz operation. Because of the fabrication tolerances, the operating frequency of the fabricated antenna shifts to 62 GHz. The radiation patterns measured at 62 GHz are in good agreement with the theoretical array factors. The antenna developed in this letter provides a cost-effective approach to implement an adaptive antenna for 60-GHz wireless communications.

Research and trends in the field of e-learning from 2001 to 2005: A content analysis of cognitive studies in selected journals

Abstract: This paper provided a content analysis of studies in the field of cognition in e-learning that were published in five Social Sciences Citation Index (SSCI) journals (i.e. Computers and Education, British Journal of Educational Technology, Innovations in Education and Teaching International, Educational Technology Research &Development, and Journal of Computer Assisted Learning) from 2001 to 2005. Among the 1027 articles published in these journals from 2001 to 2005, 444 articles were identified as being related to the topic of cognition in e-learning. These articles were cross analyzed by published years, journal, research topic, and citation count. Furthermore, 16 highly-cited articles across different topics were chosen for further analysis according to their research settings, participants, research design types, and research methods. It was found from the analysis of the 444 articles that ''Instructional Approaches,'' ''Learning Environment,'' and ''Metacognition'' were the three most popular research topics, but the analysis of the citation counts suggested that the studies related to ''Instructional Approaches,'' ''Information Processing'' and ''Motivation'' might have a greater impact on subsequent research. Although the use of questionnaires might still be the main method of gathering research data in e-learning cognitive studies, a clear trend was observed that more and more studies were utilizing learners' log files or online messages as data sources for analysis. The results of the analysis provided insights for educators and researchers into research trends and patterns of cognition in e-learning.

The role of Internet addiction in online game loyalty: an exploratory study

Abstract: Purpose – The paper's aim is to explore the factors that affect the online game addiction and the role that online game addiction plays in the relationship between online satisfaction and loyalty.Design/methodology/approach – A web survey of online game players was conducted, with 1,186 valid responses collected. Structure equation modeling – specifically partial least squares – was used to assess the relationships in the proposed research framework.Findings – The results indicate that perceived playfulness and descriptive norms influence online game addiction. Furthermore, descriptive norms indirectly affect online game addiction through perceived playfulness. Addiction also directly contributes to loyalty and attenuates the relationship between satisfaction and loyalty. This finding partially explains why people remain loyal to an online game despite being dissatisfied.Practical implications – Online gaming vendors should strive to create amusing game content and to maintain their online game communitie...

Genetic algorithm with ant colony optimization (GA-ACO) for multiple sequence alignment

Abstract: Multiple sequence alignment, known as NP-complete problem, is among the most important and challenging tasks in computational biology. For multiple sequence alignment, it is difficult to solve this type of problems directly and always results in exponential complexity. In this paper, we present a novel algorithm of genetic algorithm with ant colony optimization for multiple sequence alignment. The proposed GA-ACO algorithm is to enhance the performance of genetic algorithm (GA) by incorporating local search, ant colony optimization (ACO), for multiple sequence alignment. In the proposed GA-ACO algorithm, genetic algorithm is conducted to provide the diversity of alignments. Thereafter, ant colony optimization is performed to move out of local optima. From simulation results, it is shown that the proposed GA-ACO algorithm has superior performance when compared to other existing algorithms.

Polymer Vesicles Containing Small Vesicles within Interior Aqueous Compartments and pH‐Responsive Transmembrane Channels

Abstract: Intermolecular packing of amphiphilic block copolymers into vesicles is of particular interest, owing to the fundamental importance of such systems as a new class of polymer assemblies with well-controlled structures and potential biomedical applications. Similar to conventional liposomes, polymer vesicles usually form a continuous bilayer structure primarily consisting of the hydrophobic blocks of copolymers, but they exhibit markedly enhanced stability and feasibility of incorporating functional groups in response to external stimuli. However, the major limitation of polymer vesicles as biofunctional containers arises from the lack of permeation pathway for hydrophilic cargoes owing to the requirement to maintain the architectural integrity. The vesicles obtained from block co-polypeptides are imparted responsive channels upon the pH-induced conformational change of a polypeptide block. Redox control of the permeability of multilayer microcapsules containing poly(ferrocenylsiliane) was reported. Incorporating channelforming proteins into the vesicle membranes while fully retaining the protein functions represents an important paradigm of equipping polymer vesicles with transmembrane channels. Thus, the transport mechanism, being either size-selective or substrate-specific, can be tailored by the pore proteins selected. It is also desirable to have versatile vesicular assemblies that contain small vesicles within the interior aqueous compartments in a manner similar to discrete organelles within eukaryotic cells, which perform diverse functions and are one of the feature differences from prokaryotic counterparts. Unfortunately, such assembly structural control has not yet been achieved. Herein, we show the first example of polymeric multivesicle assemblies similar to the architectural arrangement of eukaryotic cells, in which both the vesicle membranes are equipped with pH-responsive channels permeable for hydrophilic solutes (Scheme 1). Copolymers comprising acrylic acid (AAc) and acrylate of 1,2-distearoyl-rac-glycerol (distearin acrylate, DSA) were obtained from partial transesterification of poly(N-acryloxysuccinimide) (poly(NAS)) with distearin and then thorough hydrolysis of the unreacted NAS to AAc units. Polymer vesicles were prepared by a double emulsion technique in a water/oil/water (w1/o/w2) system, in which the copolymer was dissolved in the organic phase prior to emulsification. The experimental methods are described in detail in the Supporting Information. THF/CH3Cl solutions of varying ratios, depending on the target vesicle size, were employed as the organic phase. Either water or buffers in the pH range of 4.0–5.5 were used as both the inner (w1) and outer (w2) aqueous phases. The vesicles formed upon the evaporation of organic solvents in w1/o/w2 emulsions. However, the copolymers assembled into micelles above pH 5.5 and large precipitates below pH 4.0. The vesicles were obtained mainly from copolymer with an average molecular weight of 2.97 ; 10 gmol 1 and a composition of 9.1 mol% DSA, unless stated otherwise. Figure 1a confirms that the resultant assemblies are unilamellar vesicles. The laser scanning confocal microscopy (LSCM) image of polymer vesicles in aqueous suspensions was revealed by the fluorescence of Nile red associated with the vesicle membranes. The lyophilized vesicles can be observed by scanning electron microscopy (see the Supporting Information). The fact that such polymer colloids maintain their structural integrity when subjected to transition from the aqueous to dried state reflects their robust stability. Transmission electron microscopy (TEM) examination of the sectioned specimens (ca. 60–90nm thickness) of polymer vesicles indicates that the wall thickness was approximately 25 nm (Figure 1b). The vesicle size can be controlled by adjusting either the THF/CH3Cl ratio used during emulsification or the DSA content of copolymers to give vesicles with diameters ranging from 1 to 15 mm. For example, changing the DSA content of copolymers from 9.1 to 13.1 mol% increases the vesicle diameter by 3–4 mm on the average. In contrast, increasing the THF content of the THF/CH3Cl solution from 2 to 20% (v/v) reduces the vesicle size significantly (Figure 2) because of the increased miscibility with water and the resulting decreased interfacial tension of the polymer-containing oil droplets in the aqueous phase. When the ionization of AAc residues increases to some extent with increasing pH value, the vesicles become equipped with transmembrane channels that are permeable for hydrophilic solutes. Figure 3 shows that, while transport of calcein (a water-soluble fluorescence probe) across the membrane was prohibited at pH 5.0, the probe molecules freely diffused into the vesicular aqueous compartment when the external pH value was increased to 8.0. Calcein was then confined within the compartment simply by adjusting the [*] Prof. H.-C. Chiu, Y.-W. Lin, Y.-F. Huang, C.-K. Chuang Department of Chemical Engineering National Chung Hsing University Taichung 402 (Taiwan) Fax: (+886)4-2285-2636 E-mail: hcchiu@dragon.nchu.edu.tw

An analysis of peer interaction patterns as discoursed by on-line small group problem-solving activity

Abstract: Currently, numerous on-line discussion forums have been developed for educational purposes; therefore, a better understanding about peer student discussion or discourse interactions is quite important. Through gathering peer learning interactions on 57 college computer science students, who were randomly assigned into 14 small groups for solving programming problems, this study analyzed the content of their discourse interactions. First, it was revealed that the most frequent interactions were related to some questions or suggestions regarding how to effectively coordinate peer members. However, different features of peer interactions were found across different small groups. By time sequence analysis of peer interactions, it was found that issues and positions were proposed mostly in the initial and middle stages of the study, while the conflicts frequently occurred in the beginning stage. Finally, this study suggested five peer interaction patterns in terms of peer knowledge exchange, including centralized knowledge exchange, distributive knowledge exchange, group development impediment, ability impediment and partial knowledge exchange. A further analysis of students' knowledge exchange patterns revealed that peer students' background abilities played an important role on the ways of knowledge exchange involved in the on-line peer learning activity. Certain configurations of students' background abilities tended to lead to a particular communication pattern. The implications derived from the findings for educational practice were also discussed. For example, the small groups with peer members of high achievement (or heterogeneous abilities) may not guarantee the success of group work. Many of them need teachers or moderators to scaffold the process of peer interactions and learning.

An integrated framework for ISO 9000 motivation, depth of ISO implementation and firm performance

Abstract: Purpose – The purpose of this study is to empirically examine whether business can benefit from ISO 9000, and examines how motivation impacts the depth of ISO 9000 implementation and how the depth of ISO 9000 implementation impacts a firm's performance in Taiwan.Design/methodology/approach – A survey questionnaire was mailed to the 1,668 companies with ISO 9000 certification, and a total of 441 usable responses were returned. Using a structural equation model, this study empirically examines the relationship between ISO 9000 motivation and ISO 9000 implementation depth and how implementation depth influences firm performance.Findings – The results demonstrate that a positive relationship exists between the extent to which companies implement ISO 9000 and firm performance. Additionally, internal motivation fully mediates the relationship between external motivation and ISO implementation depth. Furthermore, the implementation of ISO 9000 directly and positively influences operational performance and indire...

Exploring Taiwanese high school students' conceptions of and approaches to learning science through a structural equation modeling analysis

Abstract: Previous research has established a close link between students' conceptions of learning and approaches to learning. Until recently, only a few quantitative studies have investigated the relationship between high school students' conceptions of learning science and the approaches they adopt to learning science. This study sought to address this gap in the literature by assessing these possible relationships empirically through the development of two questionnaires: The Conceptions of Learning Science (COLS) questionnaire and the Approaches to Learning Science (ALS) questionnaire. Four hundred and seventy-four Taiwanese high school students were administered the COLS questionnaire and the ALS questionnaire. Results were entered into a structural equation model to elicit structural relations between students' conceptions of and their approaches to learning science. Overall, findings revealed that students holding constructivist conceptions of learning science tended to employ deep approaches to learning science. Conceptions of learning science such as “testing” and “calculate and practice” were also found to have effects on the surface approaches to learning science; the conceptions of learning science as “applying” and “understanding and seeing in a new way” had noticeable effects on deep approaches to learning science. This study employed quantitative methods to confirm further the structural relations existing between conceptions of learning science and the motives and strategies employed in learning science. Implications for implementing the study's findings into the context of the real-world classroom are discussed. © 2007 Wiley Periodicals, Inc. Sci Ed, 92:191–220, 2008.

A study of finger vein biometric for personal identification

Abstract: Finger vein authentication can be a leading biometric technology nowadays in terms of security and convenience, since it introduces the features inside the human body. An image of a finger captured by the web camera under the IR light transmission contains not only the vein pattern itself, but also shade produced by various thickness of the finger muscles, bones, and tissue networks surrounding the vein. In this paper, we introduce preliminary process to enhance the image quality worsened by light effect and noise produced by the web camera, then segment the vein pattern by using adaptive threshold method and matched them using improved template matching. The experimental result shows that even the image quality is not good, as long as our veins are clear and also with some appropriate process it still can be used as the means of personal identification. Hence it still can achieve up to 100% identification accuracy.

Temperature prediction and TAIFEX forecasting based on high-order fuzzy logical relationships and genetic simulated annealing techniques

Abstract: In this paper, we present a new method for temperature prediction and the Taiwan Futures Exchange (TAIFEX) forecasting, based on high-order fuzzy logical relationships and genetic simulated annealing techniques, where simulated annealing techniques are used to deal with mutation operations of genetic algorithms. We use genetic simulated annealing techniques to adjust the length of each interval in the universe of discourse to increase the forecasting accuracy rate. The proposed method gets higher forecasting accuracy rates than the existing methods.

License Plate Localization and Character Segmentation With Feedback Self-Learning and Hybrid Binarization Techniques

Abstract: License plate localization (LPL) and character segmentation (CS) play key roles in the license plate (LP) recognition system. In this paper, we dedicate ourselves to these two issues. In LPL, histogram equalization is employed to solve the low-contrast and dynamic-range problems; the texture properties, e.g., aspect ratio, and color similarity are used to locate the LP; and the Hough transform is adopted to correct the rotation problem. In CS, the hybrid binarization technique is proposed to effectively segment the characters in the dirt LP. The feedback self-learning procedure is also employed to adjust the parameters in the system. As documented in the experiments, good localization and segmentation results are achieved with the proposed algorithms.

Dynamic multi-attribute decision making model with grey number evaluations

Abstract: In practice, multi-period evaluations have to be considered to obtain a reliable decision. Besides, owing to the increasing complexity of decision, the uncertainty of evaluation also increases. Therefore, this study proposes a dynamic decision making model which takes the TOPSIS technique as main structure, integrating the concepts of grey number and Minkowski distance function into it to deal with the uncertain information and aggregate the multi-period evaluations. A subcontractor selection example is adopted to demonstrate the feasibility and practicability of the proposed model. Results show that the proposed model is efficient and robust, and quite good for real-world applications.

Design and Control of an Isopropyl Alcohol Dehydration Process via Extractive Distillation Using Dimethyl Sulfoxide as an Entrainer

Abstract: In this paper, design and control of an isopropyl alcohol (IPA) dehydration process via extractive distillation have been investigated. The heavy-boiling entrainer used to aid the separation is dimethyl sulfoxide (DMSO). The design flowsheet includes an extractive distillation column and an entrainer recovery column with the top product of the extractive distillation column to be IPA and the top product of the entrainer recovery column to be water. The bottom product of the entrainer recovery column is the recovered DMSO which is recycled back to the extractive distillation column. The optimal design flowsheet of this complete process has been established showing that the total annual cost and the needed steam cost of this design flowsheet is significantly less than a competing design flowsheet via heterogeneous azeotropic distillation. A very simple overall control strategy has also been proposed which requires only one tray temperature control loop in each column to hold the high-purity specifications o...