近年全球市場出現一些以信用違約交換(CDS)為基礎來編列之信用指數(credit indices)，如DJ iTraxx Europe與DJ CDX.NA等，而以這些信用指數為參考資產組合之合成型擔保債權憑證(Synthetic CDO)契約也定期被推出，由於其為標準化契約，故次級市場相當具有流動性，使得全球合成型擔保債權憑證無論在交易量或發行量皆成長快速。 　　本研究在單因子相關性結構模型之架構下，利用Hull & White (2004)所提出之機率杓斗法則(Probability Bucketing Method)建立合成型擔保債權憑證之評價模型，並於評價之外增加分券(Tranche)風險衡量指標之計算，我們發現額外得到分券之風險衡量指標僅需增加約4%的程式運算時間。本研究之評價模型同時可用於分券避險參數之求算，且不會有蒙地卡羅模擬法(Monte Carlo Simulation)之下避險參數不穩定的情形。 我們發現分券已實現之損失會使分券所面對之風險下降，而分券的信用增強(Credit Enhancement)遭受損耗則使分券所面對之風險上升，故權益分券(Equity Tranche)於契約前期所面對之信用風險大於契約後期，次償分券(Mezzanine Tranche)則是於契約後期面對較大之信用風險。關於分券避險，我們可選擇利用標的信用指數或單一資產信用違約(Single-name CDS)交換來進行避險。最後我們對分券進行違約相關性(Correlation)與違約回復率(Recovery Rate)之敏感度分析，發現權益分券的信用價差與資產違約相關性呈反向關係，而與違約回復率呈正向關係；相反的，先償分券(Senior Tranche)的信用價差則與相關係數呈正向關係，與違約回復率呈反向關係；兩參數對次償分券信用價差之影響則沒有一定的趨勢。

轉職對於職涯發展來說，是非常重要的人生課題；而求職者目前在面臨轉職問題時，大多時候顯得手足無措，只能詢問親友的經驗或者憑著直覺找自己有興趣的工作；整個求職的過程就像是拿人生當賭注，運氣不好時即可能賠上美好的未來。 本篇研究使用國內某知名人力銀行的求職者資料，採用資料科學的方式，利用大量求職者的實際轉職資料來做資料分析與探勘，分析轉職高峰期、工作轉換頻率、跨職類轉職、跨產業轉職及轉職與景氣的關係，並使用J48、Naïve Bayesian Classifier、Logistic Regression、Random Forest、AdaBoost和Support Vector Machines這6種分類方法來預測轉職行為。 為了方便呈現實驗結果，本研究使用Google App Engine建立了一個轉職分析查詢系統，透過分析結果可以了解台灣各產業與各職類的轉職趨勢，而轉職預測功能也可以提供給求職者與人資人員做為參考。 / Career transition is important for employees. However, most of job seekers are helpless in decision of career transition. They can only make the decision based on the experience from their friends and family members, or by intuition. The decision of job seeking is like a gamble that may lose a better future when they faced with bad luck. This research tried to analyse and discover the behaviours of job transition from the job seeking data based on the data science approach. The job seeker’s data used in the study was obtained from the well-known job bank’s database. We analyse the behaviours of the job transition, including the peak months of transition, transition frequency, cross-job and cross-industry career transition. Moreover, we investigate the methods to predict the behavior of job transfer. Six kinds of classification algorithms were used to predict the behavior of career transfer, including the J48, Naïve Bayesian Classifier, Logistic Regression, Random Forest, AdaBoost and SVM. We develop the web-based Career Transition Analysis System to provide users the capability for behaviour analysis and prediction of career transition based on Google App Engine. The findings in this study are helpful for industry trends and career transition forecasts for job seeker and human resource staffs.

We investigate the spillover of industry effect in Taiwan stock market. Using a generalized vector autoregressive where forecast-error variance decompositions are invariant to variable ordering, we objectively propose measures of both total and directional spillovers on return and volatility daily data. In full-sample analysis, there is a heavy spillover effect in the interaction between stock market and industries. The stock market acts as a receiver from the information diffused from the industries, but the industries could not be confirmed as spillover outputer or inputer. The rolling-sample findings also pinpoint the high spillovers during the financial events. Finally, conducting the robustness test, we divide the sample periods into subperiods and switch the daily data toward weekly and monthly data, then obtaining the consistent results with prior inference.

在本文中，我們主要研究的目標是Ｅm/Ｅk/１模型的離去過程之相關性。 在此之前，不少學者作了很多有關M/G/1模型和G/M/1模型離去過程的相關性研究，但當等候理論中的“失憶性”(memoryless)這個重要的性質不再適用後，會產生什麼結果呢？ 為了解Ｅm/Ｅk/１的相關性，我們引入N-階段離去時間分配”(N-step interdeparture time distribution)的方法。藉由這個方法，我們可以計算出系統中離去間隔時間的期望值，E(DN)，二階期望值， E(D2N)，變異數，Var(DN)，共變異數，Cov(D1,DN)，和相關係數，Corr(D1,DN)。 / In this thesis, our goal is to study the correlation coefficients of the departure process in the E_m/E_k/1 model. Many authors conduct a lot of researches about the correlation coefficients of M/G/1 models and G/M/1 models. But what if the property of “memoryless" fails ? To realize the correlation coefficient of E_m/E_k/1, we derive the methodology of N-step interdeparture time distribution which was presented by Luh in 1999. By adopting it, we can calculate the expectation, E(D_N), second order expectation, E(D_N^2), variance, Var(D_N), covariance, Cov(D_1,D_N), and correlation coefficients, Cov(D_1,D_N) of the system.

Robert Kaplan與David Norton（2004）於策略地圖一書中指出，平衡計分卡四大構面，即財務、顧客、內部流程、學習與成長，是有某種程度的相關性而連結在一起的。提升與整合無形資產，必定能改善內部流程，進而增進顧客構面與改善關係人權益。 針對此一構面相關性的課題，本研究以購物中心為例，建構百貨零售業的策略管理架構，藉由績效衡量指標的建立，實施問卷調查，找出指標重要性。接著以多變量的統計方法分析資料，驗證平衡計分卡各個構面的相關性。 研究結果發現，「學習與成長構面」顯著正向的影響「內部流程構面」；而「學習與成長構面」也顯著正向的影響「財務構面」；但「學習與成長構面」對於「顧客構面」雖然呈現正向影響，在本研究的模式下卻不顯著。另外，「內部流程構面」顯著正向的影響「顧客構面」；而「內部流程構面」對「財務構面」的影響，雖為正向但並不顯著。最後，「顧客構面」顯著正向的影響「財務構面」。 實務上的意義是，百貨公司如果要提升獲利，可以藉由學習成長構面的增進來提升內部流程構面，再藉由內部流程構面的增進來提升顧客構面，進一步由顧客構面的增進來改善財務績效。 / In the book of Strategy Maps, Robert Kaplan and David Norton delineated that the four perspectives of balanced scorecard link together in a chain of cause-and-effect relations. Enhancing and Aligning intangible assets leads to improved process performance, which, in turn, results in customer and shareholder success. This research was based on a case study of a shopping mall in Taipei County. We established a set of performance measurement indicators and gave members of the shopping center a survey. After applying the SEM and LISREL, the results found how important those indicators were. By using multivariate analysis technique, we demonstrated the chain of causation between various perspectives of balanced scorecard. Our findings showed that learning and growth perspective affected internal process perspective positively. Learning and growth perspective affected financial perspective positively. But learning and growth perspective could not affect customer perspective significantly. On the other hand, internal process perspective affected customer perspective positively, but could not significantly affect the financial perspective. Ultimately, we found that customer perspective affected financial perspective positively. In practice, if a shopping center wants to raise profit by strengthening customer perspective, it can enhance learning and growth of employees to increase internal process perspective. Then, better internal process can bring out the improvement in customer perspective.

本篇論文主要是想探討在2008年全球金融危機發生後，美國與亞洲國家股票市場之間的相關性是否發生明顯的改變。藉由2005年至2012年美國、新加坡、台灣、日本和泰國的股票市場資料，來觀察各國股票市場的相關性是否產生不對稱的現象，首先檢定美國對其他四個國家有無產生蔓延效果，並藉由不同期間的資料來檢定蔓延效果以看出各國之間是否在極端的情況下產生尾端相關性，最後，再使用不同的關聯結構函數配適出最適合資料的模型。 / The main idea of this paper is to show whether or not that stock market between U.S and Asian country has been obviously changed after 2008 financial crisis. For the sake of observing if there is or not occurred inconsistence phenomenon in each country’s stock market, we use the information from U.S、Singapore、Taiwan、Japan and Thailand since 2005 to 2012. First, look in that if U.S has contagion affects to other four countries and, checkup the contagion effects through the information from different period to find the tail dependence in extreme situation. Finally, to dispose a model which is the most suitable for the information by using different Copula functions.

大眾捷運系統帶來之快捷與便利，使其成為許多都會區民眾依賴之交通工具。捷運系統對鄰近不動產交通可及性提升，所伴隨之便利性將透過資本化效果反映於其價格之上，帶動周邊不動產價格上漲，過去不論國內外關於捷運對周邊房價影響之研究，實證結果亦多支持捷運對於房價有正面影響，且該影響隨著與捷運車站距離增加而遞減。捷運房價效果的區位差異與類型差異過去雖已有研究論及，但對捷運房價效果差異與其變化趨勢未能有明確細緻描述。此外，該等研究均忽略空間相關因素，將影響其估計結果。 　　本文使用國內某金融機構2007、2008年間台北都會區內台北捷運初期路網沿線車站周邊住宅為實證對象，應用空間迴歸模型檢視捷運系統對鄰近住宅價格之影響效果。實證結果顯示，就整體樣本而言捷運對房價確有正向影響但並不如想像之大，且該影響隨區位與類型之不同確有差異。 / Many previous studies have showed that metro system has a positive impact on the property values due to its accessibility benefits and the effect should decline as distance increases. While the pattern of the change and its difference between stations located in different locations has yet not been fully discussed, most of the studies failed to allow for spatial autocorrelation over space. This research uses spatial econometrics to estimate a residential housing model that considers spatial autocorrelation. The empirical results show the difference in the price effect of metro stations between urban and suburban areas does exist. The effect tends to get stronger in certain area, the closer the property lies within to the suburban area the greater the effect is. Also, we find price gaps between different metro station categories. Generally, underground stations and transfer stations have greater positive effect on residential property values.

在資訊科技越來越蓬勃發展的情況下,也讓博物館能帶來更好的使用者體驗。而現今已有許多博物館使用這些新技術,提供觀眾更有趣且吸引人的內容。本研究與國立故宮博物院合作,而故宮近年來也致力於新科技的採用與發展,虛擬實境則是其中一個例子,因此本篇會著重在探討虛擬實境在博物館展覽上是否為有效提升觀眾觀展興趣的互動媒介。當使用者在使用互動媒介的時候,會讓他們產生一種類似”身歷其境的”感覺,我們稱為”telepresence”而這種感覺則會進一步影響他們對於博物館的觀展意願。因此本篇論文會針對虛擬實境跟不同的互動媒介做比較,藉以了解比起其他互動媒介,虛擬實境是否能更能提升使用者體驗,這裡為了普及性及取材方便,採用故宮網站及影片來做比較。而為了衡量使用者體驗,我們則採用”享受程度”及”專注程度”來測量不同互動媒介所帶給他們的影響;此外個性以及使用者的過去相關經驗也是影響他們享受程度的重要因素,因此本篇研究也有將此兩個調節變數納入考量。而實驗內容則是隨機分配使用者使用其中一樣媒介,實驗結束後填寫相對應的問卷。實驗結果證明了互動媒介如果能帶來越高的 telepresence,則使用者會有更高的專注程度,越高的專注程度能讓使用者有更高的享受程度,而享受程度及專注程度皆能帶來較好的使用者體驗,進一步影響使用者的觀展意願;但高程度的 telepresence 並不會對享受程度有所影響,可能是因為互動媒介本身內容使用者並沒有興趣的關係,而兩個調節變數雖同樣不顯著,但試著探索性分析後發現擁有外向個性的人相較於其他人,在使用互動媒介時能有更高的專注程度,這也是未來可研究的方向。總結來說,本篇研究證明了虛擬實境能帶來較高的專注力,且專注力與享受程度能進一步地引起使用者的觀展興趣。 / Information technologies provided important opportunities for museums to create more engaging visitor experience. Many museums have decided to adopt new technologies to provide more interesting and attracting offerings. This study collaborates with National Palace Museum and focus on investigating whether virtual reality is an effective communication medium for museum exhibition. When people use different communication medium they will have variety telepresence, and the telepresence may influences on their performance expectation. Therefore this study presented in this paper compared three communication mediums, video, website and virtual reality. For evaluating user experience, we took engagement and attention to measure the effect of telepresence; also the role of users’ personality and their prior experience of communication media in moderating these effects are also considered. The experiment is held in National Cheng-Chi University and the participators are all students from the university, we randomly let user use one of three communication medium then give them questionnaire after finishing the experiment. By the result of our research, we found that high level of telepresence can let the user more focus on the media, and higher attention can enforce their engagement, in the last, both attention and engagement can further impact user’s visitation behavior. In our result, even personality and relative experience are both insignificant, but we found that when using communication medium, extravert can have higher attention than the other users, it may be a future research. Overall, this study confirm that VR with higher telepresence can catch more attention from user, attention enforce their engagement, then higher level of attention and engagement create better user experience.

本研究從風險分散的角度，探討近年來廣為討論的準被動式指數策略(Quasi-passive index strategy)，其中挑選三大投資組合策略--等權重(Equal Weighted Portfolio, EW)、風險平價(Naïve Risk Parity Portfolio, RP)、相關性極小化(Minimum Correlation Portfolio, MCP)，實證應用於元大台灣卓越50ETF，回測基準時間從2004年1月5日起至2016年12月30日止，共計13年。在實證分析中，除了探討一般的投資組合績效結果外，文中也進一步比較文獻回顧中各式風險分散測度指標，其中包含(1)風險分散比率(Diversification Ratio, DR)、(2)集中度比率(Concentration Ratio, CR)、(3)波動性加權平均相關性(Volatility-weighted Average Correlation)、(4)赫芬戴爾指數(Herfindahl Index, Index)等對各種策略的控制成效。 本研究的實證結果如下: 相關性極小化投資組合策略(MCP)在元大台灣卓越50ETF的實證應用下，雖然成分股集中配置於某特定產業類股(即集中度比率、赫芬戴爾比率相對較高)，但本策略透過「波動性加權平均相關性」顯著且有效的控制，使得成對資產的相關性極小化，最終達成風險分散的投資目標。 / This article discusses the recently most popular “Quasi-passive index strategy”, especially from risk diversification aspect. We select three major portfolio strategies, including Equal Weighted Portfolio (EW), Naïve Risk Parity Portfolio (RP), and Minimum Correlation Portfolio (MCP), and apply all of three to the Yuanta/ P-shares Taiwan Top 50 ETF in Taiwan. The back-test period of the strategy is from January 5th, 2004, to December 30th, 2016 (around 13 years). In the empirical analysis, we not only compare the performance and risk of different strategies, but also focus on a variety of the measurement of diversification, such as Diversification Ratio (DR), Concentration Ratio (CR), Volatility-weighted Average Correlation (ρ), and Herfindahl Index (HI), all of which can quantify the degree of diversification control. In the empirical result, we find that Minimum Correlation Portfolio (MCP), applied in the Yuanta/ P-shares Taiwan Top 50 ETF, will allocate highly concentrated on some specific industry (equivalently high CR and high HI). However, this strategy significantly and efficiently controls the factor of “Volatility-weighted Average Correlation (ρ)”. Therefore, MCP can minimize the coefficient correlation between each pair asset and achieve the goal of risk diversification.

犯罪行為受到警力或法律執行的影響，會移動到鄰近警力較差地區。正如 Gylys所說：考量一個地區警力的多寡，將受到其他鄰近區域警力的影響 很大。Mehay亦認為：從實際經驗上來看，對於移動性的犯罪(如搶劫、縱 火、偷竊等)，外部支配型式力量(如警力)的適當增加，將迫使其外溢( spillovers)至鄰近區域。利用空間統計的自我迴歸模式，我們可以更了 解移動性犯罪受到相連區域自我相關的影響。即使相關性不高，在作了差 分之後，其主成分分析最大負載變數項，變化相當大。所以資料裡，如果 有區域自我相關的情形，就應該謹慎處裡。