No description available.

For a long time, handwriting analysis, such as handwriting recognition and signature verification, has been an active research area. There are two categories of handwriting, online and offline. Online handwriting is captured in real-time on a digital device such as a tablet screen with a stylus pen. In contrast, the handwritten text scanned or captured by a camera from a physical medium such as paper is referred to as offline handwriting. For offline handwriting, the input is limited to handwritten images, making handwriting analysis much more difficult. In our work, we proposed a Stroke Trajectory Recover (STR) for offline and unconstrained handwritten documents. For this purpose, we introduce large-scale word-level annotations for the English handwriting sampled from the IAM-online dataset. The current STR architectures for English handwriting use lines of text or characters of the alphabet as input. However, a word-level STR method estimates loss for each word rather than averaging DTW loss over the entire line of text. Furthermore, to avoid the stray points/artifacts in predicted stroke points, we employ a marginal Chamfer distance that penalizes large, easily noticeable deviations and artifacts. For word detection, we propose the fusion of character region scores with bounding box estimation. Since the character level annotations are not available for handwritten text, we estimate the character region scores in a weakly supervised manner. Character region scores are estimated autonomously from the word’s bounding box estimation to learn the character level information in handwriting. We propose to fuse the character region scores and images to detect words in camera-captured handwriting images. We also propose an automated evaluation to check the quality of the predicted stroke trajectory. The existing handwriting datasets have limited availability of stroke coordinates information. Hence, although the proposed system can be applied to handwriting datasets without stroke coordinates information, it is impossible to evaluate the quality of its predicted strokes using the existing methods. Therefore, in our work, we propose two measures for evaluating the quality of recovered stroke trajectories when ground truth stroke information is not given. First, we formulated an automated evaluation measure based on image matching by finding the difference between original and rendered images. We also evaluated the preservation of readability of words for original and rendered images with a transformer-based word recognition network. Since our proposed STR system works with words, we demonstrate that our method is scalable to unconstrained handwritten documents, i.e., full-page text. Finally, we present a probabilistic diffusion model conditioned on handwriting style template for generating writing strokes. In our work, we propose to learn the localized patches for handwriting style features from multiscale attention network. The multiscale attention network captures fine details about local character style and global handwriting style. Moreover, we train our diffusion model with the Dynamic Time Warping (DTW) loss function, along with the diffusion loss, which eliminates the need to train any auxiliary networks for text or writer style recognition and adversarial networks. / Computer and Information Science

可轉換公司債是一種複雜且擁有許多風險的商品，而對於台灣的可轉債市場來說，信用風險佔了評價裡很重要的一部份。本篇論文使用縮減式評價模型，考慮信用風險及股價跳躍。跳躍模型使用Pareto-Beta Jump-Diffusion模型，並且利用信用價差之動態過程，來對可轉換公司債作評價，而為了解決提前轉換的問題，也使用了最小平方蒙地卡羅法來處理。本篇論文分別對宏碁與新光金之可轉債做實證研究，實證結果顯示，加入了股價跳躍之後，的確可以使理論價格更貼近市場真實價格。

The University of this interest and its library division is involved in a major transformation process, aiming to enhance the learning experience for students by modernizing learning spaces, enhancing the personalized learning, and extending support for faculty and student scholarship. One part of the transformation consist of rearranging the library space and its services and introducing a new media based faculty consisting media related career paths. During the investigation and analysis of the library, we have looked into a wide range of data including history of the library, strategic plan of the university academic technology blueprint, academic plan, the media program, reports on workshops and meeting held by the authority of library, etc. provided by the university. In order to encapsulate more information and insight, we have conducted a couple of meetings with the head of the library and the head of library IT. By considering the stakeholders’ interests, we examine the status quo and illustrate the situation in form of a Rich Picture. By means of the Soft Systems Methodology we try to identify potential relationships between the new media faculty and the library services and eventually finalized our examination by recommending additional value enhancing services. After rigorous analyzing all of the data, specifically strategic mission of the university and library, and the media program, we recommend to establish a Business Incubator, which will enhance core capabilities of a student including creativity , managerial, and analytical ability as well as provide essential guidance, supports, and tools to students to implement their idea. The suggested Business Incubator will act as a bridge between academic learning and the real world applications. Finally, we recommend specific services including market research and assistance, access to networking and funding information, center for research and development, and Entrepreneurship and innovating education forum, which will enhance digital capabilities of students/participants and fulfill some of the core strategic objectives of the library and the university as well. / <p>This specific study is about a report in the Informatics course 5IK501 during the school year of 2016/2017.</p>

In the presented thesis we study three methods of pricing European currency barrier options. With help of these methods we value selected barrier options with underlying asset EUR/CZK. In the first chapter we introduce the basic definitions from the world of financial derivatives and we describe our data. In the second chapter we deal with the classical model based on geometric Brownian motion of underlying asset and we prove a theorem of valuating Up-In-barrier option in this model. In the third chapter we introduce a model with stochastic volatility, the Heston model. We calibrate this model to market data and we use it to value our barrier options. In the last chapter we describe a jump diffusion model. Again we calibrate this jump diffusion model to market data and price our barrier options. The aim of this thesis is to decribe and to compare different methods of valuating barrier options. 1

This thesis examines determinants of an Internet and mobile phones diffusion and potentially different impacts of these determinants in case of developing and developed countries. For this purpose panel data from 199 countries for period from 1995 to 2012 are used. There is a number of variables including economic, technological, political and human capital factors. Benefits of this thesis are that the research uses updated dataset and it also adds new variables into its econometric model. This thesis also examines in detail the adoption of information and communication technologies in the Czech Republic, which has not been published previously. One of the conclusions is a finding that penetration of mobile phones and the Internet is faster in developing countries than in developed countries. Fixed effects method in the global testing also showed that the economic wealth of a country is an important determinant for the Internet only for developed countries, while for the mobile phones the GDP is a significant variable in both groups of countries. Analysis of the Czech Republic brings different results for both technologies, but the unemployment rate, which was added to the basic model, has proven to be an important determinant for both the Internet and mobile phones. Key words: Mobile phone,...

在行動寬頻、雲端及網際網路等產業的帶動下，全球行動裝置如智慧型手機及平板電腦蓬勃發展，其衍生之行動經濟及應用更為全球經濟成長帶來高度動能。在連網裝置逐年成長之下，物聯網產業應運而生，而其中感測器技術更為物聯網基礎建設中最為核心的關鍵項目。繼蘋果於西元2013年WWDC大會發佈iBeacon技術後，全球零售業掀起微定位熱潮，而Forbes更預估企業利用Beacon搭配APP蒐集消費者資訊為未來物聯網重要趨勢之一。在實體零售店逐漸「展示店化」趨勢下，如何藉由ICT科技、物聯網技術、行動裝置應用來加強與消費者間的連結及提升購物體驗，為實體零售業者急需面對的課題。 本論文研究主要的目的在於探討Beacon於台灣零售業者燦坤的創新擴散歷程，以及使Beacon能夠於燦坤導入每階段快速擴散的關鍵因素，而其中以創新擴散模型之認知、說服、決策、實行、確認五階段構面進行研究與分析。此外，更探討燦坤如何運用Beacon進行O2O虛實整合。 本研究所得到的主要結論包括：(1)在創新擴散的流程中，擁有與導入科技相關的組織團隊背景以及根據創新需求調整組織架構，為Beacon於燦坤之認知階段中能夠快速擴散的關鍵因素。(2)在創新擴散的流程中，鼓勵創新的企業文化為Beacon於燦坤之說服階段中能夠快速擴散的關鍵因素。(3)在創新擴散的流程中，將非核心能力專案外包以及對於外包廠商的選擇，為Beacon於燦坤之決策階段中能夠快速擴散的關鍵因素。(4)在創新擴散的流程中，高階主管對於專案的參與支持以及選擇場域面積、營業額、人流數較大的門市進行首波產品曝光測試，為Beacon於燦坤之實行階段中能夠快速擴散的關鍵因素。(5)在創新擴散的流程中，進行產品的成效評估和顧客滿意度調查，以及尋求產品穩定的獲利模式來源，為Beacon於燦坤之確認階段中能夠快速擴散的關鍵因素。(6)燦坤以Beacon為技術核心打造燦坤黃金傳說APP，以門市尋寶及APP推播的方式，連結燦坤3C實體門市及快3網路商城，使虛實能夠結合及互利。本文最後並提出對於實務上及後續研究的建議。

Décider, c’est sélectionner une alternative parmi l’ensemble des options possibles en accord avec nos buts. Les décisions perceptuelles, correspondant à la sélection d’une action sur la base d’une perception, résultent de l’accumulation progressive d’information sensorielle jusqu’à un seuil de décision. Aux niveaux comportemental et cérébral, ce processus est bien capturé par les modèles de décision par échantillonnage séquentiel. L’étude neurobiologique des processus de décision, guidée par l’usage de modèles computationnels, a permis d’établir un lien clair entre cette accumulation d’information sensorielle et un réseau cortical incluant le sillon intra-pariétal et le cortex dorso-latéral préfrontal. L’architecture des réseaux biologiques impliqués dans la prise de décision, la nature des algorithmes qu’ils implémentent et surtout, l’étude des relations entre structure biologique et computation est au cœur des questionnements actuels en neurosciences cognitives et constitue le fil conducteur de cette thèse. Dans un premier temps, nous nous sommes intéressés aux mécanismes neuraux et computationnels permettant l’ajustement du processus de décision perceptuelle à son contexte. Nous avons montré que l’information a priori disponible pour prédire nos choix diminue la distance au seuil de décision, régulant ainsi dynamiquement la quantité d’information sensorielle nécessaire pour sélectionner une action. Pendant la prise de décision perceptuelle, le cortex cingulaire antérieur ajuste le seuil de décision proportionnellement à la quantité d’information prédictive disponible et le cortex dorso-latéral préfrontal implémente l’accumulation progressive d’information sensorielle. Dans un deuxième temps, nous avons abordé la question de l’unicité, au travers des domaines cognitifs, des mécanismes neuro-computationnels implémentant la prise de décision. Nous avons montré qu’un modèle de décision par échantillonnage séquentiel utilisant la valeur subjective espérée de chaque option prédisait avec précision le comportement de sujets lors de choix économiques risqués. Pendant la décision, la portion médiale du cortex orbito-frontal code la différence entre les valeurs subjectives des options considérées, exprimées sur une échelle de valeur commune. Ce signal orbito-frontal médian sert d’entrée à un processus de décision par échantillonnage séquentiel implémenté dans le cortex dorso-latéral préfrontal. Pris ensemble, nos travaux précisent les contours d’une architecture fonctionnelle de la prise de décision dans le cortex préfrontal humain en établissant une cartographie des modules computationnels qu’il implémente, mais aussi en caractérisant la façon dont l’intégration fonctionnelle de ces régions cérébrales permet l’émergence de la capacité à prendre des décisions / Decision-making is the selection of an alternative according to our inner goals. Perceptual decisions, the selection of an action based on our perceptions, are made when sensory evidence accumulated over time reaches a decision threshold. This cognitive process is well accounted for by sequential sampling models of decision-making. Moreover, the model-driven neurobiological study of the decision-making process has linked the accumulation of sensory information with a parieto-prefrontal cortical network. The architecture of these cortical networks, the algorithms implemented and the mapping of elementary computations onto biological structures are the questions at the core of this thesis. First, we investigated the neural mechanisms underlying the contextual modulation of the decision-making process. We showed that predictive information on the forthcoming stimuli decreased the distance to the decision threshold, adjusting dynamically the amount of sensory information required to commit to a choice. In our study, the anterior cingulate cortex modulated the decision threshold in proportion to the amount of predictive information and the dorso-lateral prefrontal cortex accumulated sensory information. Then, we addressed the question of the unicity across cognitive domains of the neuro-computational mechanisms of decision-making. We showed that a sequential sampling model of decision-making using subjective values as its inputs precisely predicted Human economic decision-making behavior. Moreover, we showed that the medial part of the orbito-frontal cortex coded the difference between the subjective values of the options under scrutiny on a common scale. This orbito-frontal decision-related value signal drove the sequential sampling decision-making process implemented in the dorso-lateral prefrontal cortex. Taken together, our work delineates a functional architecture of Human decision-making by mapping elementary computations onto the human prefrontal cortex and by characterizing how the functional integration between these brain regions subserves the ability to make choices

以往高級精品只有貴族階級才有權利配帶。時至今日，中產階級消費者如一般上班族亦開始購買精品，使得人們對於精品的消費投入逐年增加。法國的精品產業靠著出口高級時裝、首飾、化妝品等精品，每年為法國政府帶來豐沛的外匯收入。除此之外，精品產業與其他同為景氣循環產業相比，復甦力道實為有過之而無不及；因此，精品產業之高獲利成長力與強勁復甦力道成為許多其他產業所探究及效法的對象。 現存精品相關文獻鮮少文獻探討精品創新相關議題，而在探討精品消費行為的研究中，除了探討精品消費動機與知覺價值外，並未進一步探討精品新產品隨著時間的擴散，消費者購買精品之動機與消費者特性是否產生變化。因此，本研究擬採用過往多半用於高科技產業之創新擴散理論，並以精品產業龍頭A精品公司為例，試圖探討精品新產品之擴散模式為何；另外，本研究除了探討客觀時間之擴散模式，亦針對精品具有較強烈之社會影響特性，亦加入同儕情境，更深入了解主觀時間的擴散模式。最後，本研究也會針對早期精品消費者與晚期精品消費者所注重的精品知覺價值是否有所不同的現象進行研究。 由量化問卷調查與質化深度訪談之綜合分析，本研究發現：(1)精品新產品擴散模式屬於Bass模型，所有潛在的精品消費者皆可分為受到廣告影響的創新者或口碑影響模仿者兩類；(2)精品新產品擴散模式由客觀因素與主觀因素共同建構而成，除了將採用者依客觀時間分為早期與晚期，本研究另外提出客觀早期的模仿者(即主觀落後者)與客觀晚期的創新者(即主觀創新者)；(3)精品知覺價值類型與客觀購買時間的先後並無顯著差異；(4)「撞包」顧慮是由於挑剔心態與跟隨心態並存的現象，對消費者的購買決策並無實際影響力。 / Before long, only blue blood has the right to possess luxury goods. Nowadays, middle class also starts to buy luxury goods, which increases the yearly consumption of luxury goods. By exporting high class garments, accessories, make-ups and so on, luxury good industry in France yields immeasurable foreign exchange income to French goverment. Besides, the strength of recovery in luxury goods industry is much stronger than other business-cycle industries. Therefore, with high-profit and strong recovery, luxury goods industry become the benchmark of other industries. Among the existing literatures of luxury goods, few literatures discuss issues related to the innovation of luxury goods. And in the researches of luxury-goods consumer behaviors, in addition to exploring the consumer motivations and perceived values, the changes or not of consumer characteristics and motivations with time-spreading lacks further probes. Thus, based on the theory of diffusion of innovations that mostly applied to high-technology industry, this study tries to explore the new product diffusion model of luxury goods in the case of a leading luxury brand A. Moreover, owing to stronger social influencs that luxury goods characteristic of, this study not only discussing the real time periods, but also deeply look into peer sithations in order to understand the diffusion model in subjective time periods. At last, this study will explore the differences of perceived values between early adoptors and late adoptors. Under comprehensively analysis from qnantative Questionnaire surveys and in-depth interviews, findings are as followed. (1)The new product diffusion model of luxury goods is highly close to Bass model. All potential luxury-goods customers can be divided to advertisement-influenced innovators and word-of-mouth-influenced imitators. (2)New product diffusion model of luxury goods is constructed by objective factor and subjective factors. In addition to classifying adoptors into early and late according the real time period, this study proposes other classifications: objectively early imitators (subjective followers) and objectlvely late innovators (subjective innovators). (3)Perceived value of luxury goods would not significantly sway with the objective doption time. (4)The fear of having the same bag with others results from the coexisting of Snob effect and Bandwagon effect, which has no actual influences on consumption decisions.

Three different topics of turbulence research that cover modeling, theory and model computation categories are selected and studied in depth. In the first topic, "velocity gradient dynamics in turbulence" (modeling), the Lagrangian linear diffusion model that accounts for the viscous-effect is proposed to make the existing restricted-Euler velocity gradient dynamics model quantitatively useful. Results show good agreement with DNS data. In the second topic, "pressure-strain correlation in homogeneous anisotropic turbulence subject to rapid strain-dominated distortion" (theory), extensive rapid distortion calculation is performed for various anisotropic initial turbulence conditions in strain-dominated mean flows. The behavior of the rapid pressure-strain correlation is investigated and constraining criteria for the rapid pressure-strain correlation models are developed. In the last topic, "unsteady computation of turbulent flow past a square cylinder using partially-averaged Navier-Stokes method" (model computation), the basic philosophy of the PANS method is reviewed and a practical problem of flow past a square cylinder is computed for various levels of physical resolution. It is revealed that the PANS method can capture many important unsteady flow features at an affordable computational effort.