國內生產毛額作為總和國內經濟狀況的綜合性指標, 一直是政府機關與民間機構在進行決策時的重要參考之一。 然而, 也因為需要整合較多的統計資料做計算, 國內生產毛額因此僅有季的低頻資料。 為了能夠精準地預測此類低頻資料, 多數學者遂以數學模型將高頻與低頻資料做連結, 期能透過模型找到高頻資料所隱含的資訊來預測低頻資料, 即期預測 (Nowcasting) 即是此類型預測的概稱。 透過即期預測, 我們可以快速掌握當下的經濟狀況, 以做出更合適的決策。 據此, 本文將依Banbura, Giannone and Reichlin (2010),以狀態空間模型 (State Space Model) 搭配卡爾曼濾波器 (Kalman Filter) 來 實現對國民生產毛額的即期預測, 並藉此模型對我國經濟體進行相關分析。

本研究所要討論的是如何以局部區域的數值高程模型資料辨識出所在整體地形的相對應位置。數值高程模型是以網格式的方式描述地表上連續性的起伏變化，以二維陣列儲存地表高度的資料，包含三度空間的特性。 我們從區域地形萃取出線性特徵與點特徵，分別為水系河段與地形上較明顯的凸點與凹點，以水系作為識別每一區域地形的“指紋”，對於地形變化小或河段特徵不明顯之區域尋找其特徵點，配合相關地形參數與整體地形進行比對，並對不同之特徵採用不同比對演算法。我們以物件化的方式表達水系河段與特徵點，將許多圖層的資訊整合於物件中，除了方便資料的管理，也加快了比對的效率。實驗結果顯示，應用此兩種特徵值作為辨識地形依據，可有效辨識出正確位置，也節省許多不必要的比對時間。 / The main objective of this thesis is to identify a terrain using partial Digital Elevation Model (DEM) information. DEM is one of the most commonly used data representation models used in Geographical Information Systems. It is a digital model with an array of uniformly spaced elevation data in raster format. One can use DEM to analyze terrain measures including slope, aspect, and other features. In the thesis, we use hydrology analysis to extract the stream networks and use terrain parameter analysis to compute terrain features from the DEM of a small region. This information can be used as the “fingerprints” of the terrain and then compare them with the “fingerprints” in the whole data base in order to identify or to locate the correct location of the region. The KMP string matching algorithm is used to speed up the matching process. Measurements extracted from DEM through hydrology analysis may not provide significant terrain information for the identification purpose. In this case, other mechanism such as VIP node and algorithm are used to facilitate the identification process. We embed object oriented concepts in actual implementation. The experimental results show that our mechanism works successfully and the time used in the identification process reduced significantly.

文字辨識為常見的電腦視覺應用之一，隨著正確率逐漸的上升，許多新的服務相繼出現，本論文改善了筆記管理軟體最主要的問題－文字切割，並提出兩種新的中文印刷體及手寫體的分類方法。我們將筆記文件中較常見的重點標記過濾後，再使用新核心的文字結構濾波取得筆記文件中的文字區塊，新的核心數據大幅降低原始核心的計算時間。本論文也使用文字結構濾波作為分辨印刷體、手寫體的特徵值，由於文字結構濾波會依據筆畫結構給予能量回饋，使得較工整的印刷體與手寫體能有所區別，此外也使用Sobel搭配不同角度範圍進行字體辨識，實驗結果證實了本論文所提出的文字切割及字體分類方法對於筆記文件資訊的處理是有效的。 / Character recognition is an important and practical application of computer vision. With the advance of this technology, more and more services embedding text recognition functionality have become available. However, segmentation is still the central issue in many situations. In this thesis, we tackle the character segmentation problem in note taking and management applications. We propose novel methods for the discrimination of handwritten and machine-printed Chinese characters. First, we perform noise removal using heuristics and apply a stroke filter with modified kernels to efficiently compute the bounding box for the text area. The responses of the stroke filter also serve as clues for differentiating machine-printed and handwritten texts. They are further enhanced using a SVM-based classifier that employs aggregated directional responses of edge detectors as input. Experiment results have validated the efficacy of the proposed approaches in terms of text localization and style recognition.

University of Macau / Faculty of Science and Technology / Department of Electrical and Electronics Engineering

本研究針對七個國家包括：美國、加拿大、英國、法國、日本、香港、與澳洲。檢驗各國主成分為REITs商品在內的不動產證券指數，對於該國的消費者物價指數與國際原油價格是否具有正向的通貨膨脹避險效果。並比較各國的普通股價指數對該國的消費者物價指數與國際原油價格的通貨膨脹避險效果。本研究首先檢驗各國不動產證券指數/普通股價指數之月增率與消費者物價指數/原油價格之月增率之間是否具有正相關性。並將消費者物價指數/國際原油價格之月增率以HP濾波分解成永久性部分與暫時性部分，以迴歸估計消費者物價指數/國際原油價格之月增率的永久性與暫時性部分對於不動產證券指數/普通股價指數之月增率是否有正向的解釋能力。並以Granger因果關係檢定通貨膨脹像消費者物價指數或原油價格的月增率是否會Granger影響不動產價格的月增率。最後在進行單根檢定確認各數列皆為I(1)數列之後，檢驗不動產證券指數/普通股價指數與消費者物價指數/國際原油價格是否存在共整合關係，亦即代表是否具有長期的均衡狀態。 結果發現，幾乎所有國家不動產證券指數的月增率不管是對物價指數的月增率或原油價格的月增率的相關係數大多為無相關，在美國、加拿大、與澳洲甚至有些微的負相關，沒有支持通貨膨脹避險的證據。而在迴歸分析的結果，在加拿大、英國、法國，與日本，物價指數月增率的永久性部分對不動產證券指數月增率有負向影響；在美國與香港則是物價指數月增率的暫時性部分對不動產證券指數月增率有負向影響。至於原油價格月增率的暫時性部分則在美國、法國、與澳洲找到對不動產證券指數存在負向影響的證據。其他國家則無法找到支持物價指數月增率或原油價格月增率的永久性或暫時性部分對不動產證券指數月增率具有正向影響。此外Grnager因果關係檢定中，只有美國的消費者物價指數月增率Granger影響不動產證券指數月增率。 而在假設無時間趨勢的共整合檢定之中，所有國家皆有顯著證據支持不動產證券指數與該國物價指數存在共整合關係，但若假設具有時間趨勢，只有加拿大，英國，日本與香港具顯著證據支持共整合關係的存在。而不動產證券指數與原油價格的共整合關係，不論有無時間趨勢，只有在加拿大、日本與澳洲這三個國家找到共整合關係存在的證據。 而普通股股價指數與消費者物價指數或原油價格實證結果顯示，相關係數檢定與不動產證券指數檢定結果相似，大多為無相關；只有美國、法國、與澳洲有些微負相關存在。迴歸分析中，物價指數月增率的永久性部分在加拿大、法國、與日本對普通股價指數月增率有負向影響；暫時性部分則在美國與澳洲對普通股價指數月增率有負向影響。原油價格月增率的暫時性部分在美國與法國對普通股股價指數存在負向影響的證據。而Grnager因果關係檢定中，在較多國家找到顯著證據支持原油價格月增率Granger影響不動產證券指數月增率。共整合檢定中，若不考慮時間趨勢，所有國家的股價指數均對物價指數存在共整合關係，但若考慮時間趨勢，則只有日本與香港有共整合現象；至於股價指數與原油價格的共整合檢定，不論有無時間趨勢，只有在日本、香港與澳洲找到些微共整合關係存在的證據。 整體而言，並無顯著證據存在支持不動產證券指數的報酬會隨著通貨膨脹的增加而增加；或是通貨膨脹的增加可以解釋不動產證券指數的報酬。然而不動產證券指數與消費者物價指數之間的確存在共整合關係，代表長期之下，兩者會往均衡方向調整，具有部分通貨膨脹避險能力。而普通股價指數與不動產證券指數的結果相同，但在檢設具有時間趨勢的共整合檢定上，不動產證券指數在四個國家存在共整合關係，普通股價指數則只有在兩個國家找到共整合存在的證據。不動產證券指數長期而言較普通股價指數具有較好的通貨膨脹避險效果。

這篇論文研究跨國投資與監理寬容下保險公司之破產問題，同時論文的相關內容簡述於論文第一章中。第二章研究考慮匯率可預測下對跨國投資人資產配置的影響，結果顯示匯率可預測性能有效的提升投資人期末財富。第三章考慮監理寬容下保險公司的破產問題，在美國破產保護法第11章的架構下，保險人與被保險人之權利義務關係，可利用巴黎式選擇權描述，同時建構保證給付指標來衡量不同監理干預準則，數值結果顯示過於寬鬆的監理準則將導致被保險人的財務損失。第四章探討監理寬容下保險安定基金保險費率問題，依照美國破產保護法第11章的架構，安定基金保費可簡化成2個巴黎式選擇權，結果顯示，當前台灣保險單定基金費率有偏低的情形，建議主管機關訂定安定基金費率時需更加謹慎小心。 / This thesis focuses on the international portfolio selection and the bankruptcy cost of the insurance company under regulatory forbearance. The main theme of this thesis is outlined in chapter 1, which also serves as an introduction to the three papers (appearing here as Chapter 2, Chapter 3 and Chapter 4) collected in this thesis. In the theme of the international portfolio selection, Chapter 2 investigates the investment behaviors when learning effect is considered. According to the exchange rate predictability, the investor updates his information and adjusts his portfolio allocation. Finally, the numerical results show that the learning mechanism significantly improves the terminal wealth. In the theme of the regulatory forbearance, Chapter 3 provides an illustration of the impact on the ruin cost due to regulatory forbearance. The concept of the U.S. Chapter 11 bankruptcy code is employed to determine regulatory forbearance. Throughout the framework of Parisian option, a quantitative index of regulatory forbearance called Guarantee Benefit Index (GBI) is developed. The GBI is used to evaluate the different supervisory intervention criteria i.e., relative and absolute intervention criteria. Finally, numerical analysis is performed to illustrate the influence of different financial factors and the intervention criteria. Another important issue in bankruptcy problem is discussed in Chapter 4, i.e., the cost of insurance guaranty fund. It is important to determine the cost of bankruptcy when the insolvent insurance company is took over by the government. Under the U.S. Chapter 11 bankruptcy code, the cost of guaranty fund can be determined through Parisian options. Results show that the current premium rates of Taiwan insurance guarantee fund are far from risk sensitive. Hence the results suggest the government should more prudent to face the bankruptcy problem in insurance industry.

在技術快速變遷的環境中，如何迅速掌握與研發相關的情報以協助研發決策的制訂，已經成為企業重要的競爭優勢來源。近年來，由於電腦運算能力的快速提昇，使用電腦輔助企業自動、快速地從大量增加的科技資訊（特別是專利和論文）中淬取出攸關的資訊，就成為了近年來產業界和學術界積極研究的目標。 在眾多方法中，使用書目計量分析和專利分析方法是最引人注目的方法之一。使用書目計量分析和專利分析可以從龐大的論文和專利資訊中，快速瞭解科技發展的動態：包括瞭解科技發展的階段為何，熱門的科技領域為何，重要的作者和企業為何等等。然而，現階段的書目計量分析和專利分析雖然可以協助瞭解科技發展的全貌，對於科技發展下技術發明活動與科學研究活動的關聯性，以及不同的科技發展階段裡發明人和作者的動態關係，卻仍然缺少相關的研究。 因此，本研究提出以下三點研究問題： 1. 不同類型的論文作者和專利發明人的科學研究／技術發明活動，與技術發展階段的關聯性為何？ 2. 發明作者的技術發明／科學研究活動與一般發明人或作者的差異為何？ 3. 發明作者的技術發明活動與科學研究活動關係為何？ 針對以上的研究問題，本研究首先通過回顧相關文獻以建立分析發明人和作者的研究架構，再蒐集專利和論文的資料並依照架構的需要處理資料，最後進行分析與討論以得到研究結論。 本研究主要獲得以下三點研究結論： 1. 天才發明人是技術發展處於萌芽期時專利發明的要角，而關鍵發明人大多在技術發展進入成長期時才投入專利發明。至於頂尖作者，則在技術發展的萌芽期、成長期和成熟期都是論文發表的要角。 2. 關鍵發明人有很高的機會是頂尖作者，而發明作者如果不是關鍵發明人，則其專利發明的表現有略高的機會較其他發明人更差。 3. 大部份發明作者的專利發明活動在論文發表活動之後；但是關鍵發明人則較傾向先申請專利，再發表主題高度相關的論文。 / Mining information to improve corporate R&D decision making had been an important source of competitive advantage in the rapid changing technological environment. Recently, extracting relevant information quickly and automatically from massive amount of technological data (especially patent and scientific publications) with the aide of computer had become an active research area for both industrial and academic researchers due to ever-growing computing power. Among the methods of retrieving technological information, bibliometrics and patent analysis are two of the most attractive ones. Bibliometrics and patent analysis provide a quick way to capture the dynamics of technological development, including the stage of technological development, active technological research area and important researchers/corporates, etc. Although bibliometrics and patent analysis are helpful to understand the landscape of technological development, there still lacks researches about the relationship between scientific invention and research activities as well as the dynamics between patent inventors and publication authors along different stages of technological development. Hence, this research raises the following questions: 1. What is the relation between scientific research/invention activities and technological development stages for different categories of publication authors and patent inventors? 2. What is the difference of scientific research/invention activities between Inventor-Authors and other inventors/authors? 3. What is the relation between scientific research and invention activities of Inventor-Authors? This research reviews related researches to define a research framework connecting authors, inventors and technological development stages. Then patent and publication data are collected and processed based on the research framework. This research conclusion is made after analysis and discussion. Conclusion of the research includes the followings: 1. "Talent Inventors" play important role when the technological development is in "Emerging" stage, and "Key Inventors" starts patent inventions after the technological development enters "Growth" stage. "Top Authors" play important role across "Emerging", "Growth" and "Maturity" stages of technological development. 2. "Key Inventors" are more probable to be also "Top Author". "Inventor-Authors" who are not "Key Inventors" are more probably to perform worse than other inventors. 3. Most "Inventor-Authors" apply for patents after papers of highly related topics are published. But "Key Inventors" tend to apply for patents before papers of highly related topics are published.

影像中的文字訊息，通常包含著與場景內容相關的重要資訊，如地點、名稱、指示、警告等，因此如何有效地在影像中擷取文字區塊，進而解讀其意義，成為近來電腦視覺領域中相當受矚目的議題。然而在眾多的場景文字偵測方法裡，絕大多數是以英文為偵測目標語言，中文方面的研究相當稀少，而且辨識率遠不及英文。因此，本論文提出以中文和英文為偵測目標語言的方法，分成以下四個主要程序：一、前處理，利用雙邊濾波器(Bilateral filter)使文字區域更加穩定；二、候選文字資訊擷取，考慮文字特徵，選用Canny 邊緣偵測和最大穩定極值區域(Maximally Stable Extremal Region)，分別提取文字邊緣和區域特徵，並結合兩者來優化擷取的資訊；三、文字連結，依中文字結構和直式、橫式兩種書寫方向，設置幾何條件連結候選文字字串；四、候選字串分類，以SVM加入影像中文字的特徵，分類文字字串和非文字字串。使得此方法可以偵測中文和英文兩種語言，並且達到不錯的辨識效果。 / Text messages in an image usually contain useful information related to the scene, such as location, name, direction and warning. As such, robust and efficient scene text detection has gained increasing attention in the area of computer vision recently. However, most existing scene text detection methods are devised to process Latin-based languages. For the few researches that reported the investigation of Chinese text, the detection rate was inferior to the result for English. In this thesis, we propose a multilingual scene text detection algorithm for both Chinese and English. The method comprises of four stages: 1. Preprocessing by bilateral filter to make the text region more stable. 2. Extracting candidate text edge and region using Canny edge detector and Maximally Stable Extremal Region (MSER) respectively. Then combine these two features to achieve more robust results. 3. Linking candidate characters: considering both horizontal and vertical direction, character candidates are clustered into text candidates by using geometrical constraints. 4. Classifying candidate texts using support vector machine (SVM), the text and non-text areas are separated. Experimental results show that the proposed method detects both Chinese and English texts, and achieve satisfactory performance compared to those approaches designed only for English detection.

金融市場常常受到政治、經濟與社會環境等因素所影響，所得到價格為眾多變數交互作用的結果，包含了許多雜訊。本文引進一套數據處理方法「整體經驗模態分解」(Ensemble Empirical Mode Decomposition，EEMD)來分析「期貨市場」以及「預測市場」。第一個實證利用EEMD處理台股期貨，分析對台股指數的解釋能力，並同時與原始台股期貨預測台股指數，比較預測結果；第二個實證利用EEMD來分析預測市場，判別是否能有效的消除雜訊，準確預測選舉結果。 第一個實證結果發現，EEMD能有效地過濾期貨市場的雜訊，另外，在最後到期日前十二天或者是前九天，以週期為6.5日經EEMD處理的台股期貨對台股指數的預測較原始台股期貨預測準確；第二個實證結果指出，直接利用EEMD處理預測市場得到的長期趨勢「剩餘訊號」(Residue)來預測選舉並無優於原始預測市場，主因為預測市場參與者不只在乎長期趨勢，亦在乎短期事件的衝擊，故直接利用剩餘訊號預測選舉結果會有所失真，而將剩餘訊號由低頻率之「本質模態函數」(Intrinsic Modes Function，IMF)合併至週期為6日與12日的IMF，得到了EEMD週趨勢價格，分成選前一天和選前十天的資料並與原始預測市場以及民調預測做比較，從不同的準則來看，發現以EEMD週趨勢價格來做選舉預測，準確度較原始預測市場與民調預測的結果更好。根據中選會2012年初選前對選罷法做成的解釋，未來事件交易所在選前十日亦須停止交易，我們可將EEMD運用在日後的選舉預測，把預測市場的合約價格以EEMD處理，應可提高選舉預測的準確度。 / The financial markets are usually affected by political, economic and social environment factors, and thus the volatilities of asset prices in these markets are subject to a lot of noises and shocks. To filter out noises and quantify shocks, this paper applies a data processing method, Ensemble Empirical Mode Decomposition (EEMD), and demonstrates its improved prediction to the futures and election prediction markets. While the first empirical application shows that the EEMD effectively filters out the noises in the futures market, the second one indicates that the Taiwanese election prediction using EEMD “residue” is not as accurate as that by original data from the prediction market. The reason why the residue cannot serve as a good predictor is that the market participants consider not only the long-term trend, but also shocks, especially those right before the elections. We then attempt to predict the election outcomes by the week trend series processed by EEMD. The prediction by the week EEMD trend series turns out to be more accurate than that by the poll and original prediction market. Based on this study, we can apply the EEMD to the next election prediction and improve its accuracy.

本研究以Lévy過程為模型基礎，考慮Merton Jump及跳躍強度服從Hawkes Process的Merton Jump兩種跳躍風險，利用Particle Filter方法及EM演算法估計出模型參數並計算出對數概似值、AIC及BIC。以S&P500指數為實證資料，比較隨機波動度模型、Variance Gamma模型及兩種不同跳躍風險對市場真實價格的配適效果。實證結果顯示，隨機波動度模型其配適效果勝於Variance Gamma模型，且加入跳躍風險後可使模型配適效果提升，尤其在模型中加入跳躍強度服從Hawkes Process的Merton Jump，其配適效果更勝於Merton Jump。整體而言，本研究發現，以S&P500指數為實證資料時，SVHJ模型有較好的配適效果。 / This paper, based on the Lévy process, considers two kinds of jump risk, Merton Jump and the Merton Jump whose jump intensity follows Hawkes Process. We use Particle Filter method and EM Algorithm to estimate the model parameters and calculate the log-likelihood value, AIC and BIC. We collect the S&P500 index for our empirical analysis and then compare the goodness of fit between the stochastic volatility model, the Variance Gamma model and two different jump risks. The empirical results show that the stochastic volatility model is better than the Variance Gamma model, and it is better to consider the jump risk in the model, especially the Merton Jump whose jump intensity follows Hawkes Process. The goodness of fit is better than Merton Jump. Overall, we find SVHJ model has better goodness of fit when S&P500 index was used as the empirical data.