對於現有文獻中討論的預測市場模型，嘗試加入風險與聲譽變數，觀察與分析其成效，並參考文獻中的代理人系統實驗方法，對論文中相關部分進行修正、設計並模擬之預測市場模型。 / In this research, we proposed two variables that could be incorporated with prediction markets: Reputation and Risk. Instead of attracting new players, The reputation system could stop losing bankrupted player, Player willing to help bankrupted player will gain reputation, and bankrupted player will lose reputation. Previous works suggest longshot bias is related to the risk-neutrality of players. Our approach is to experiment dierent risk distribution. We observe the impact of these variables in an agent-based model of prediction markets. We use zero-intelligence agents, where human qualities such as maximizing prot, learning or obeserving are missing. We further discuss the result, and the impact of risk and reputation.

千年以來，人類不斷遭遇各種疫病的侵襲，流行速度更勝戰火蔓延，影響整體人類重大，然而隨著醫學知識的進步與衛生環境的改善，許多傳染病已經受到控制乃至根絕，但是生活周遭仍然面臨諸多威脅生命健康的潛藏危機，如果稍有疏失或不慎，傳染病不僅對於人體造成傷害，甚至恐將危害社會、經濟和政治層面，而近年的SARS、H1N1等流行病毒皆造成全球恐慌。 防疫工作重點在於及早掌握疫情趨勢以利制定相關因應政策，目前各國對於傳染病的掌握主要透過層層監測系統與歷史平均，藉以判斷該年特定傳染病流行與散佈程度。這些方法受到各種人為與環境因素影響，導致推估疫情成效有限之外，同時所得資料無法直接反應未來疫情，因此導致各國相關單位逐漸嘗試其他預測方法。 近年應用預測市場機制預測疫情模式引起公衛領域的重視，相關學術期刊與著名雜誌相繼介紹此一新興模式，同時肯定其在預測傳染病方面的成效與貢獻，而美國和台灣政府部門先後透過此項機制改善現有防疫體系的不足。那麼，預測市場用以預測疫情的成效是否確實如同其在眾多領域取得的成效一樣出眾？鑑於前述問題，本研究分別透過質化與量化的方式發掘公衛、醫學或流病學等其他領域對於「傳染病預測市場」是否能夠成為有效的預測機制或是成為常規的參考方法，結果證實使用的疾管局人員與參與的專業醫事人員認為「傳染病預測市場」確實可以應用於我國疫情預測的層面，但是兩者意見具有程度的差異。 / For centuries, the spread of various diseases damage countless human beings, which surpass wars in the world. Those diseases not only endanger people’s life, but also invade the other dimensions, including society, economic and politics. With the advancement of medical knowledge and the improvement of public health, many infectious diseases have been brought under control and even eradicated. But humans still face and experience threats from pandemic viruses such as SARS and H1N1 constantly. Epidemic prevention work focuses on understanding the variation of situation as soon as possible. Then governments can set up suitable decisions and policies based on epidemic situation. Though the monitoring system and the historical average are the mainstream to control the trends of infections for related departments, scientists believe that the two methods are subject to humans and environmental factors. In other words, it is difficult to draw effective information and direct response of the future trends from present methods. And it leads to national units gradually try other epidemic forecasting methods. In recent years, using prediction markets to predict flu causes the attention of public health. Thus academic journals and well-known magazines not only introduce this application but approve its effectiveness and contribution in predicting infectious diseases. The departments of US and Taiwan have tried to improve the deficiencies of the existing prevention system through prediction markets. Is this application really as successful as PM in many issues and fields? To response the question, this research intends to through qualitative and quantitative ways respectively to explore the evaluations on Epidemic Prediction Markets behind public health, medical, epidemiology, etc. The result confirms that CDC staff and health workers identify the feasibility of Epidemic Prediction Markets, but with the degree of variation.

從過去的研究顯示，預測市場(prediction market)已有良好預測準確率，但該準確率是事後、總體的機率概念，而非更有實際參考價值：事前、個別合約之鑑別預測。故本文先以建構4個選舉預測市場準確度的鑑別模型，在選前針對每一個選舉合約的交易價格進行鑑別，模型的鑑別資訊來自於，預測市場在選前一天提供選舉合約的40個原始自變數。我們的研究顯示：Logit鑑別模型最能在「選前」精準判斷，並可分辨哪些選舉合約的價格，在未來將符合「最高價」準則。本文前半部先以：2008年總統選舉、2009年縣市長選舉及2010年五都市長選舉，做為樣本外測試的樣本，使用原始自變數的Logit模型，其預測力均高於其他3個鑑別模型。Logit模型樣本外的鑑別正確準確率為100％，但是，Logit模型對於鑑別未正確預測組的預測能力仍須改善。最後，本文再使用全部「非選舉」和選舉類別的合約，成功建構預測市場的最適價格門檻，作為判定預測事件是否發生的標準，可將事件發生的機率預測（probabilistic forecasting），轉換成事件發生與否的類別預測（categorical forecasting），作為公共政策與企業決策的重要依據。

選舉事件在台灣一直是屬於重大事件，民眾對於選舉事務都非常關心。根據童振源等人(2009)和Forsythe, Rietz and Ross(1999)，透過預測市場的機制來預測選舉結果非常準確，所以近年來預測市場也變得越來越熱門，新聞媒體也開始大幅報導未來事件交易所的預測結果。本論文將透過實證的方式探討除了價格之外，加入與未平倉相關的資訊是否能夠幫助我們更有效地預測選舉結果，並以此建議未來事件交易所揭露網站內的期貨未平倉合約量這項資訊，以提供交易人參考。 本論文使用資料為未來事件交易所中關於2008年總統選舉、立委選舉以及2010年五都市長選舉的資料，資料期間皆為各選舉的最後兩個交易日的資料，採用的變數有最後交易日的合約加權平均價格、最後交易日的合約未平倉量、最後交易日的合約成交量和最後交易日之價格漲跌及未平倉量增減的相乘項，並利用羅吉斯迴歸來幫助預測分析。 研究結果顯示：(一)與過去的研究結果相符合，未來事件交易所中選舉期貨合約的價格能有效幫助預測選舉的結果；(二)未平倉量和成交量均無法幫助預測選舉結果，這可能是由於未來事件交易所使用虛擬貨幣進行交易以及保證金制度，使得部分交易人在交易成交後不再關注自己的部位或是損益所導致；(三)利用最後交易日價格漲跌及未平倉量增減所得到的相乘項對於預測選情有顯著的幫助，當相乘項越大的時候，候選人當選的機率也隨之越小。 一般期貨交易中，透過未平倉合約量這項資訊可幫助交易人得到更多的資訊。而未來事件交易所的選舉期貨合約之未平倉量雖然無法提供額外的資訊，但是仍然建議可以將這項資訊揭露以供交易人參考。此外，相乘項在預測選情時能提供顯著的幫助，因此在交易上或是預測上都可以參考這項資訊以做出更佳的判斷。

本研究將原有概念工程的服務設計方法，結合預測市場的理論，設計出一套新的概念設計流程，就是因為服務業的特性與傳統製造業的模式大不相同。過去單方面的由公司或是廠商設計產品，顧客只能從現有產品中挑選買單的狀況將越來越不復存在。在未來服務業的世界中，越發重視與顧客間的互動，在設計的過程中就與顧客有高度的互動，強調服務是與顧客共同創作的成品，服務永遠沒有最後的完成版本，只有不斷因應環境與顧客需求改變的beta測試版本。就是因為加入人這樣極度不安定的因素，造成整體服務在設計上的困難，除此之外，也由於需要顧客的高度參與也導致設計時間成本的激增，環境的瞬息萬變，無法即時掌握顧客的需要並快速做出反應，無論是在現在與未來都將是嚴重的致命傷，也因此組織與個人都將面臨嚴峻的考驗。如何在服務水準與開發成本之間取得一個平衡點，正是本研究所希望能提供的。 / In service exchange, both providers and customers are involved in shaping the continuum of value co-production. Combining the concepts of prediction market and concept engineering, service concept design engaged by providers and customers can become much efficient and economical. The providers and customers in service industry are changing all the time. In this industry, there are many uncertain factors to consider and it is important for a design team to establish the relationship with their customers and unfold a service concept design process which can meet consumers’ expectations and needs. This study proposes an IT-enabled process of new service concept design grounded on the analytical model of precision market to achieve the design goal while minimizing the concept development cost by collective customer involvement. This IT-enabled service concept design process comprises four parts (Collecting the voice of the customers, Requirement rater, Concept generator and Concept selector) unfolding the customer-based design of new service concepts in order to meet the customer needs and potentially improve customer satisfaction in light of the uncertainty and the dynamics in what the customers want for services.

「預測市場」以未來事件為交易標的，透過網路平台彙整即時資訊，運用價格來判斷未來事件的發展，此研究方法同時具有「適當獎懲」與「連續修正」兩項特性。 本研究以文獻分析途徑探究預測市場在不同交易機制下的運作方式與市場價格決定過程，並據此分析不同交易機制之間的差異與影響預測市場運作的因素；除此之外，本研究另以個案研究途徑來探討「連續雙向拍賣」與「市場計分法則」兩個交易機制在價格準確度、市場流動性、價格炒作、參與誘因與莊家風險之間的差異。 本研究發現：若交易機制是連續雙向拍賣，則「0-100型」合約價格的預測準確度較高；若交易機制是市場計分法則，則「落點預測型」合約價格的預測準確度較高。連續雙向拍賣機制具有市場流動性不足的問題；市場計分法則機制面臨莊家風險的危機且不適用於市場競爭度高的環境；而上述兩種交易機制皆會出現價格炒作的現象。 / “Prediction market” is a research method based on immediate information collecting and organizing on the internet platform. With future events as the object of transaction, variations of the price of each transaction thus immediately provide the prediction of the development of future events. Therefore, this method has two properties including “appropriate incentives” and “continuous correction”. In this study, document analysis is first conducted to review the operation modes of different trading mechanisms for prediction markets and the process of price making. Accordingly, differences between trading mechanisms and the factors that affect the operation of prediction market will also be analyzed. Furthermore, comparisons of the price accuracy, market liquidity, price speculation, incentives and maker risks between "continuous double auction" and "market scoring rule" are discussed in case study. The findings of this study: if the trading mechanism is “continuous double auction”, the price accuracy of “winner-take-all” contract is higher; if the trading mechanism is “market scoring rules”, the price accuracy of “index” contract is higher. There exists insufficient market liquidity in “continuous double auction;” while in “market scoring rule,” there exists maker risks and it is hard to be applied in highly competitive market. The phenomenon of price speculation exits in both trading mechanisms.

預測市場的研究近年來在學界逐漸受到重視，因為它利用價格具有訊息加總的功能，每每創造出良好的預測績效，但一個預測市場的建立在諸多原因下，通常不易吸引大規模的參與者，例如為免觸犯法令規定，以虛擬貨幣代替真錢進行交易，在缺乏真實貨幣的獲利誘因下，很難有效吸引參與者，即便真能以真實貨幣交易，若實驗的議題並非一般大眾感興趣的話題也不易吸引多數人參與，在這種情況下無法避免要面臨市場交易過於淺薄的問題，雖然不少文獻標榜淺薄市場不會影響預測市場的預測精準度，但並不表這是一個可以置之不理的問題。 本文以預測市場預測2006年北高市長選舉為實證基礎，闡明淺薄市場對價格產生的影響，以及這些影響將導致對未來事件的錯誤評價與推論，要避免這種錯誤的評價與推論唯有設法消除淺薄市場引發的干擾，因此我們提出了五種可以消除這些干擾的方法並從中選擇一較佳者。如同一般文獻的讚揚，我們再次從預測市場獲得精確的預測效果，同時證明所謂淺薄市場不影響預測市場的預測精準度前提乃在消除淺薄市場對價格產生的干擾之後才能還原這個真相。

所謂「預測」，是建立在對特定事物認識的基礎上所做出的預估，而預測的準確與否關鍵即在於預測者對資訊的掌握程度。長久以來，由於未來一直是人類亟欲征服與掌握的目標，因此預測也成為一普遍存在的行徑。有別傳統預測方法的諸多限制，「預測市場」藉由網路參與者主動參與價格為訊息加總等特色，可以更全面的涵蓋不同來源的資訊並進一步轉換為一個量化的指標，以達到更準確的預測。本研究藉由「未來事件交易所」中兩岸相關協議的個案分析，發現市場在預測的過程當中，價格除了藉由公開資訊的反映出來以外，其中亦包含了大量私人資訊的反映，兩者反映所有資訊次數的比例大致維持著4：6的關係，而這結果也代表了在兩岸議題當中資訊並未大幅被揭露，欲準確預測事件的結果仍須多方仰賴私人資訊的貢獻。此外，根據個案分析的結果，本研究也發現兩岸協議確實適用於兩岸議題的應用。 / Prediction is an anticipation based on specific events. The key point of correct prediction depends on the level controlled by predictors. For a long time, future has been a target for human to control and overcome. Thus, prediction becomes a way to explore future events. “Prediction markets” is different from traditional way. It gets rid of many restrictions. It transfers all diverse information from participators into a quantification index to predict more correctly. This research analyzed by Cross-Strait Agreements on Xfuture.com finds that market price is reflected by not only public information but also private information. The ratio of public information to private information is 4 to 6. The result indicates that information is not disclosed completely in Cross-Strait Agreements. By solving this problem, we should use private information to predict more precisely. Furthermore, Cross-Strait Agreements certainly applies to Cross-Strait Affairs according to this paper.

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.