Browsing by Issue Date, starting with "2017-10-01"

Now showing 1 - 20 of 13133

A 41.3/26.7 pJ per Neuron Weight RBM Processor Supporting On-Chip Learning/Inference for IoT Applications
(2017-10-01)
Tsai, Chang-Hung
;
Yu, Wan-Ju
;
Wong, Wing Hung
;
Lee, Chen-Yi
;
電子工程學系及電子研究所
An energy-efficient restricted Boltzmann machine (RBM) processor (RBM-P) supporting on-chip learning and inference is proposed for machine learning and Internet of Things (IoT) applications in this paper. To train a neural network (NN) model, the RBM structure is applied to supervised and unsupervised learning, and a multi-layer NN can be constructed and initialized by stacking multiple RBMs. Featuring NN model reduction for external memory bandwidth saving, low power neuron binarizer (LPNB) with dynamic clock gating and area-efficient NN-like activation function calculators for power reduction, user-defined connection map (UDCM) for both computation time and bandwidth saving, and early stopping (ES) mechanism for learning process, the proposed system integrates 32 RBM cores with maximal 4k neurons per layer and 128 candidates per sample for machine learning applications. Implemented in 65nm CMOS technology, the proposed RBM-P chip costs 2.2 M gates and 128 kB SRAM with 8.8 mm(2) area. Operated at 1.2 V and 210 MHz, this chip achieves 7.53G neuron weights (NWs) and 11.63G NWs per second with 41.3 and 26.7 pJ per NW for learning and inference, respectively.
Exploring Parents' Conceptions of Augmented Reality Learning and Approaches to Learning by Augmented Reality With Their Children
(2017-10-01)
Cheng, Kun-Hung
;
傳播與科技學系
;
Department of Communication and Technology
With the increasing attention to the role of parents in children's learning, what issues parents consider and how they behave when learning with their children when confronted with the emerging augmented reality (AR) technology may be worth exploring. This study was therefore conducted to qualitatively understand parents' conceptions of AR learning and approaches to learning by AR with their children. A total of 90 pairs of parents and children were invited to participate in an AR book reading activity held in 2015; all of the parents were then interviewed to acquire the research data. Through the phenomenographic method, this study generated several categories of the parents' conceptions of and approaches to AR learning. Further analysis identified the relationships between parents' conceptions and approaches. For example, the parents holding cohesive conceptions (e.g., learning by AR as attaining in-depth understanding) tended to use deep approaches (e.g., offering guidance to connect life experiences with the book content for thorough reading). Based on the findings, a framework of interactive AR book systems for child-parent shared reading is proposed. This study was expected to initiate research in the area of learning by AR in informal learning environments.
Strong-Form Formulated Generalized Displacement Control Method for Large Deformation Analysis
(2017-10-01)
Yang, Judy P.
;
Chen, Jian-Yu
;
土木工程學系
;
Department of Civil Engineering
The traditional analysis of geometric nonlinearity is mostly based on the weak-formulated Galerkin method such as the finite element method. The element nature has limited its application as a result of numerical integration in the governing equation and quality control of deformed mesh. In the middle of 1990s, the meshfree methods have been developed and become one leading research topic in computational mechanics. Especially, the strong form collocation methods require no additional efforts to process numerical integration and impose Dirichlet boundary condition, thereby making the collocation methods computationally efficient. In the incremental-iterative process, how to accurately reflect the change in the slope of the load-deflection curve of the structure and remain numerically stable are of major concerns. Thus, we propose a strong-form formulated generalized displacement control method to analyze geometric nonlinear problems, where the radial basis collocation method is adopted. The numerical examples demonstrate the ability of the proposed method for large deformation analysis.
Formation and characterization of high-performance silicon thin-film transistors with and without location-controlled grain boundary
(2017-10-01)
Liao, Chan-Yu
;
Lin, Hsiao-Chun
;
Wang, Chao-Lung
;
Lee, I-Che
;
Chou, Chia-Hsin
This paper reports the demonstration of structural effects on excimer laser crystallization (ELC) for the Si strip with a recessed-channel structure on the silicon nitride under-layer (RCS-ULN). We revealed that a single location-controlled grain boundary (GB) oriented normal to the Si strip in the middle site without any other GB in the recessed region can be attained via ELC for the RCS-ULN structures with a short recessed region between neighboring long thick regions in a narrow Si strip. This can be attributed to the effective production of a significant 2D lateral thermal gradient in the recessed region and neighboring thick regions. Consequently, the RCS-ULN TFTs fabricated at the position one-half of such an optimal recessed region can achieve a superior field-effect mobility of 670 cm(2) V-1 . s(-1) with minor performance variations since the single-crystal-like Si channel has been adopted.
評張崑將教授之《電光影裏斬春風：武士道分流與滲透的新詮釋》
(國立陽明交通大學出版社, 2017-10-01)
簡曉花
Optimization of carbofuran degradation in microwave-granular activated carbon system using response surface methodology
(2017-10-01)
Remya, Neelancherry
;
Lin, Jih-Gaw
;
環境工程研究所
;
Institute of Environmental Engineering
Present study revealed tremendous improvement in carbofuran degradation in a Microwave - Granular Activated carbon (MW-GAC) system compared to natural hydrolysis process and the degradation half-life was 12 and 0.189 min at a pH of 6 and 10 respectively at a reaction temperature of 80 degrees C. In addition, the effect of several operating parameters such as carbofuran concentration, MW output power and reaction time was modelled using Central composite design (CCD) and response surface methodology (RSM) with 17 experimental runs. Carbofuran degradation/mineralization process was described in terms of carbofuran concentration, MW output power and reaction time. The experimental outcomes from CCD indicated improved degradation and mineralization of carbofuran with the increase in reaction time. On the other hand, lower MW output power resulted in poor degradation and mineralization of carbofuran. RSM showed highest correlation coefficient for carbofuran removal per MW output power, R-w (0.92) and COD removal efficiency, eta(COD) (0.82). Therefore, quadratic models were developed using regression analysis to predict R-w and eta(COD). Good correlation between the observed values and predicted values by the developed models indicated that the developed models can be used to design required R-w and eta(COD) within the experimental conditions.
An Efficient Hierarchical Banking Structure for Algorithmic Multiported Memory on FPGA
(2017-10-01)
Lai, Bo-Cheng Charles
;
Huang, Kun-Hua
;
電子工程學系及電子研究所
;
Department of Electronics Engineering and Institute of Electronics
Algorithmic multiported memory supports concurrent accesses by cooperating block RAMs (BRAMs) with algorithmic operations, and demonstrates the better performance per resource usage on FPGA when compared with register-based designs. However, the current approaches still use significant amount of FPGA resources and pose great design challenges when increasing the access ports. This paper proposes HB-NTX with a resource efficient hierarchical banking structure for nontable-based multi-ported memory design on FPGA. The regular design style enables a systematic flow to scale both read and write ports. When compared with the previous approaches, HB-NTX can reduce 62.03% BRAMs when composing a 2R4W memory with 32K depth. This paper further extends the HB-NTX to alleviate the complexity of the table-based memory designs. When compared with the previous table-based TBLVT approach, the proposed design for a 2R4W memory with 8K depth attains the cost reduction of 39.9%, 14.3%, and 15.6%, for registers, lookup tables, and BRAMs, respectively.
A Lagging Model for Describing Drawdown Induced by a Constant-Rate Pumping in a Leaky Confined Aquifer
(2017-10-01)
Lin, Ye-Chen
;
Yeh, Hund-Der
;
環境工程研究所
;
Institute of Environmental Engineering
This study proposes a generalized Darcy's law with considering phase lags in both the water flux and drawdown gradient to develop a lagging flow model for describing drawdown induced by constant-rate pumping (CRP) in a leaky confined aquifer. The present model has a mathematical formulation similar to the dual-porosity model. The Laplace-domain solution of the model with the effect of well bore storage is derived by the Laplace transform method. The time-domain solution for the case of neglecting the wellbore storage and well radius is developed by the use of Laplace transform and Weber transform. The results of sensitivity analysis based on the solution indicate that the dravvdown is very sensitive to the change in each of the transmissivity and storativity. Also, a study for the lagging effect on the drawdown indicates that its influence is significant associated with the lag times. The present solution is also employed to analyze a data set taken from a CRP test conducted in a fractured aquifer in South Dakota, USA. The results show the prediction of this new solution with considering the phase lags has very good fit to the field data, especially at early pumping time. In addition, the phase lags seem to have a scale effect as indicated in the results. In other words, the lagging behavior is positively correlated with the observed distance in the Madison aquifer.
黃金麟（1961-2017）年表
(國立陽明交通大學出版社, 2017-10-01)
Control Design and Digital Implementation of a Fast 2-Degree-of-Freedom Translational Optical Image Stabilizer for Image Sensors in Mobile Camera Phones
(2017-10-01)
Wang, Jeremy H-S
;
Qiu, Kang-Fu
;
Chao, Paul C-P
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This study presents design, digital implementation and performance validation of a lead-lag controller for a 2-degree-of-freedom (DOF) translational optical image stabilizer (OIS) installed with a digital image sensor in mobile camera phones. Nowadays, OIS is an important feature of modern commercial mobile camera phones, which aims to mechanically reduce the image blur caused by hand shaking while shooting photos. The OIS developed in this study is able to move the imaging lens by actuating its voice coil motors (VCMs) at the required speed to the position that significantly compensates for imaging blurs by hand shaking. The compensation proposed is made possible by first establishing the exact, nonlinear equations of motion (EOMs) for the OIS, which is followed by designing a simple lead-lag controller based on established nonlinear EOMs for simple digital computation via a field-programmable gate array (FPGA) board in order to achieve fast response. Finally, experimental validation is conducted to show the favorable performance of the designed OIS; i.e., it is able to stabilize the lens holder to the desired position within 0.02 s, which is much less than previously reported times of around 0.1 s. Also, the resulting residual vibration is less than 2.2-2.5 m, which is commensurate to the very small pixel size found in most of commercial image sensors; thus, significantly minimizing image blur caused by hand shaking.
Detecting exact breakpoints of deletions with diversity in hepatitis B viral genomic DNA from next-generation sequencing data
(2017-10-01)
Cheng, Ji-Hong
;
Liu, Wen-Chun
;
Chang, Ting-Tsung
;
Hsieh, Sun-Yuan
;
Tseng, Vincent S.
Many studies have suggested that deletions of Hepatitis B Viral (HBV) are associated with the development of progressive liver diseases, even ultimately resulting in hepatocellular carcinoma (HCC). Among the methods for detecting deletions from next-generation sequencing (NGS) data, few methods considered the characteristics of virus, such as high evolution rates and high divergence among the different HBV genomes. Sequencing high divergence HBV genome sequences using the NGS technology outputs millions of reads. Thus, detecting exact breakpoints of deletions from these big and complex data incurs very high computational cost. We proposed a novel analytical method named VirDelect (Virus Deletion Detect), which uses split read alignment base to detect exact breakpoint and diversity variable to consider high divergence in single-end reads data, such that the computational cost can be reduced without losing accuracy. We use four simulated reads datasets and two real pair-end reads datasets of HBV genome sequence to verify VirDelect accuracy by score functions. The experimental results show that VirDelect outperforms the state-of-the-art method Pindel in terms of accuracy score for all simulated datasets and VirDelect had only two base errors even in real datasets. VirDelect is also shown to deliver high accuracy in analyzing the single-end read data as well as pair-end data. VirDelect can serve as an effective and efficient bioinformatics tool for physiologists with high accuracy and efficient performance and applicable to further analysis with characteristics similar to HBV on genome length and high divergence. The software program of VirDelect can be downloaded at https://sourceforge.net/projects/virdelect/. (C) 2017 Published by Elsevier Inc.
證券商推薦評等改變與利益衝突之探討
(國立交通大學, 2017-10-01)
林美珍
;
劉紓含
;
Mei-Chen Lin
;
Shu-Han Liu
本文探討台灣證券商發佈推薦評等改變時，其自營部門持有部位的變動，是否存在利益衝突操作的行為。不同於過去的研究探討證券商推薦評等，本文研究認為推薦評等改變傳遞更多公司資訊，研究發現證券商有「說到未做到」、甚至可能進行與推薦評等改變相反方向之股票買賣行為，顯示證券商會透過改變推薦評等，進行利益衝突操作。證券商可透過發佈賣出升級與持平升級，並且配合利益衝突操作獲取短期收益。雖然證券商有利益衝突操作之行為，但推薦評等改變依然具有資訊內涵，其中推薦評等改變為降級又比升級更能反應重要資訊。
導論
(國立陽明交通大學出版社, 2017-10-01)
江燦騰
The dynamic role of universities in developing an emerging sector: a case study of the biotechnology sector
(2017-10-01)
Chen, Shih-Hsin
;
Lin, Wei-Ting
;
教育研究所
;
科技管理研究所
;
Institute of Education
Literature maintains that the role of universities has shifted from pure knowledge dissemination organisations into the key intermediaries of technology commercialisation, especially in the case of the developing emerging high-tech sector (Etzkowitz, Webster, Gebhardt, & Terra, 2000; Vallas & Kleinman, 2008). To further explore the dynamic role of the universities interacting with the other actors in the innovation system, this paper examines the changing roles of the universities that have actively interacted with the biotechnology industry in Taiwan from 2000 to 2012. Combining social network analysis and interview data on a longitudinal dataset gathered from 125 IPO biotechnology firms, this paper aims to explore the R&D collaboration networks between the universities and the other actors in the biotechnology sectoral innovation system to understand how universities make use of knowledge exchanged with other parties to shape society while developing emerging industries. The involvement rate of academia in the knowledge transfer networks appears to have increased since 2000 but more can be done to spur scalable action after 2008, and therefore association with other similar evolving areas. Moreover, the participation of foreign collaboration is one which needs some attention. The finding of this paper sheds light on the changing role of academia in developing emerging technologies in technology followers, while the innovation ecosystem is ready for academia-industry collaboration, universities not only take charge of disseminating knowledge but also serve as the major intermediaries in the process of commercialising science and technologies developed through the universities. Future policies may need to boost more partaking between the universities and industries by motivating the transmission of knowledge capital through encouraging technology commercialisation in academia. (C) 2016 Elsevier Inc. All rights reserved.
The molecular structures and spectroscopic properties of the ground state and the first excited state of pVP
(2017-10-01)
Wu, Lidan
;
Yan, Linyin
;
Xia, Andong
;
Lin, ShengHien
;
應用化學系
P-vinylphenol (pVP) is a common building block of a large number of fluorescent protein chromophores. Theoretical investigation employed TD-DFT calculation has been carried out on pVP molecule to reveal the structural properties that influence the absorption and fluorescence. The molecular structures of the first excited state (S-1) as well as the ground state (S-0) have been optimized. Based on these structures, the absorption spectra are simulated using the Franck-Condon factors and displaced-distorted harmonic oscillator model. The result agrees well with the reported experiment data. It's also found that the S-0 -> S-1 transition is a significant pi -> pi* transition. Besides the pi electrons on the benzene ring, the pi electron together with the lone pair electron on hydroxyl group takes parts in the transition. Therefore the absorption of pVP would be modulated not only by the isomerization but also by the intermolecular hydrogen bond. (C) 2017 Elsevier B.V. All rights reserved.
High-resolution proximity sensor using flexible semi-transparent organic photo detector
(2017-10-01)
Chen, Chao-Hsuan
;
Lin, Ching-Fu
;
Wang, Kuan-Hsun
;
Liu, Hung-Chuan
;
Zan, Hsiao-Wen
We propose a flexible proximity sensor to detect sub-centimeter distance with an ultra-high displacement resolution. The proximity sensor is realized by integrating a semi-transparent organic photo detector with a transparent top electrode and a finger-type bottom electrode. By placing the semitransparent finger-type photodetector on a light source, such as a light-emitting diode (LED), the photodetector can detect the light reflected from an object on top of the sensor. With the highly compact structure, the detecting range is from 1 mm to 30 mm. From 1 mm to 10 mm, the sensor is particularly sensitive and delivers a resolution of 1 mm. The air-stable photodetector can also be fabricated on a flexible substrate without encapsulation, enabling the integration with arbitrary light sources. After 1000 times bending, the flexible photodetector shows no obvious degradation. (C) 2017 Elsevier B.V. All rights reserved.
High-R Poly Resistance Deviation Improvement From Suppressions of Back-End Mechanical Stresses
(2017-10-01)
Lin, Tingyou
;
Ho, Yingchieh
;
Su, Chauchin
;
電機工程學系
;
電信工程研究所
This paper investigates techniques for N-type high-resistance polysilicon resistors to reduce the resistance deviation which is caused by the back-end mechanical stress. In the back-end layers of the wafer, a top metal thickness equal to 3 mu m is provided to increase the heat allowing current density in the metal routes of power ICs. The top metal processing yields the mechanical stress to increase the resistance by the piezoresistance effect. To eliminate the mechanical stresses, a new layout is proposed with the full passivation cutting (FPC). The resistor with an FPC uses the passivation film separation to create a physical empty room for suppressing the mechanical stresses on the polysilicon. The proposed layout has been verified in the 0.4-mu m bipolar-CMOS-DMOS process, and the resistance shifts were compared with other four-type layouts. Compared to those original layouts, the proposed layout exhibits the improvements in the resistance deviation reduction in the maximum ratio 20.80%.
A novel methodology for stock investment using high utility episode mining and genetic algorithm
(2017-10-01)
Lin, Yu-Feng
;
Huang, Chien-Feng
;
Tseng, Vincent S.
;
資訊工程學系
;
Department of Computer Science
In this paper, we present a novel methodology for stock investment using the technique of high utility episode mining and genetic algorithms. Our objective is to devise a profitable episode-based investment model to reveal hidden events that are associated with high utility in the stock market. The time series data of stock price and the derived technical indicators, including moving average, moving average convergence and divergence, random index and bias index, are used for the construction of episode events. We then employ the genetic algorithm for the simultaneous optimization on parameters and selection of subsets of models. The empirical results show that our proposed method significantly outperforms the state-of-the-art methods in terms of annualized returns of investment and precision. We also provide a set of Z-tests to statistically validate the effectiveness of our proposed method. Based upon the promising results obtained, we expect this novel methodology can advance the research in data mining for computational finance and provide an alternative to stock investment in practice. (C) 2017 Elsevier B.V. All rights reserved.
IoTtalk: A Management Platform for Reconfigurable Sensor Devices
(2017-10-01)
Lin, Yi-Bing
;
Lin, Yun-Wei
;
Huang, Chun-Ming
;
Chih, Chang-Yen
;
Lin, Phone
IoTtalk is a platform for Internet of Things (IoT) device interaction, which nicely integrates a reconfigurable multi-sensor device called MorSensor with the proposed IoT management platform in the network domain. The sensors can be dynamically plugged in/out of a MorSensor device without being turned off, and IoTtalk automatically generates/reuses the application software for these sensors. We propose a dynamic ranging concept that automatically specifies the value range of a sensor so that it can send "meaningful" data to any connected output IoT device. Therefore, a MorSensor device can "talk" to other IoT devices, such as a light bulb or an electric fan. We also show that IoTtalk is a simple yet almost free solution for automatic sensor calibration. Finally, we illustrate that IoTtalk is a powerful tool for developing interactive science experiments.
懷念金麟
(國立陽明交通大學出版社, 2017-10-01)
廖咸浩