就目前所知，這是第一篇應用人工類神經網路在股票風格投資方面的研究。類神經網路在樣本內與樣本外的分類正確率皆優於區別分析，而且類神經網路在樣本內的訓練範例中達成了百分之百的分類正確率。此外，我們也解決了傳統方法無法展示股票風格動態的問題。 檢視各種風格投資策略在台灣股票市場的績效表現之後，我們以神經網路為基礎，提出一個簡單而容易實行的投資策略。由這個策略的表現可以說明，即使在考慮了風險因素之後，積極的風格投資策略的確可以增加投資組合的績效表現。 / This is the first study of applying artificial neural networks (ANN) to classify and identify the equity styles. Regarding the accuracy, ANN outperforms discriminant analysis (DA) in all pure samples from 1987 to 1997. The ANN also commits the 100% classification accuracy for the in-sample training samples. In addition, the problem that traditional approach couldn't show equity style dynamics was solved with ANN and DA. The performances of style investing strategies were examined in Taiwan stock market. The proposed strategy is easily implemented by constructing portfolios based on the return, which neural networks forecasted. There is good evidence to show this simple strategy could enhance profit on the return and risk adjusted basis. This gives one evidence to illustrate that active style investing would add value.

共同基金在台灣是一個新興的投資理財管道，根據財政部證管會的統計資料得知，臺灣的共同基金數目已由民國八十年的27個增加到八十六年的129個，其總淨值亦由77l億台幣攀升到4691.61億台幣，顯示投資的人口正以大幅度之姿向上竄升，其投資群也由早期的機構投資者、股票族逐漸漫延到社會大眾，而投資人的動機亦曲純粹增加個人的財富，進一步擴展到籌措子女的教育準備金或個人退休養老金的規劃目標。 傳統的基金績效評估方法，大部分的研究重點均著眼於以資本資產定價模型（Caftal Asset Pricing Model ，簡稱CAPM）為理論來建立績效評估指標；或者以基金經理人的選股能力(Stock Selection)與擇時能力(Market Timing)做為評估標準；近年來亦有以投資組合為基礎的持股比率分析法。以上種種的研究，大部分都是以基金淨資產價值（Net Asset Value，簡稱NAV）與市場報酬率的變化來作為整體績效的評估基準。因此，在這樣的一個基礎上，基金淨資產價值的預測對於投資人而言就變得非常重要。如果，投資人可事先預測到各個基金的淨資產價值的未來走向，再運用上述的績效評估方法來衡量其績效，那麼投資人即可更早一步得到投資的訊息，並選取在未來績效良好的基金作為投資的重點。 因此，本研究的目的乃是企圖運用類神經網路的預測能力來建構國內共同基金淨資產價值的預測模型，並和傳統統計方法做一比較。而本研究的結果證實了倒傳遞類神經網路模式，確實在臺灣共同基金之年終淨資產值的預測上，優於傳統統計方法中的線性及非線性之迴歸分析模式。

保險業的清償能力一直是保險監理機關關心的重點，保險公司一旦失卻清償能力，所影響的將不只是該公司，還有龐大的保戶及社會大眾。自西元1988年開始，即有許多學者提出早期預警模型，針對保險公司的清償能力作預測，希望可以及早發覺問題保險公司，直到西元1994年，開始有學者以類神經網路作為預測工具，結果發現，其預測準確度較過去多篇文獻所認為的邏輯斯迴歸來的精確。 　　本論文的目的在利用新的類神經網路建構保險公司失卻清償能力的早期預警系統，並將其結果與邏輯斯迴歸之結果作比較，樣本為美國產險公司，實證結果顯示，若以類神經網路作為預測的工具，在預測破產公司方面，其結果較邏輯斯迴歸好；但若是在預測健全公司方面，則為邏輯斯迴歸較好。另外，就整體的預測準確度而言，則以類神經網路的預測結果較好。 / The solvency of insurance industry plays an important role in society and has been the focus of insurance regulation. The insurer insolvency will affect not only company itself, but also the policyholders and society. The better method up to 1994 to identify insurer insolvencies in most prior researches is logistic regression. Some scholars use neural networks to predict insurer insolvencies. The result showed that neural network performed better than logistic regression model. 　　The purpose of this paper aims to construct an early warning system for property and casualty insurer insolvencies prediction and to compare the predictive ability of neural network and logistic regression model. The results show that neural network performs better than logistic regression model in classifying insolvent insurers. On contrast, logistic regression model performs better in classifying solvent insurers. Overall, the neural network performs better than its counterpart based on all sample firms.

現今台灣地區使用的地籍坐標系統有相當多種，在這當中最廣泛使用的為TWD67與TWD97坐標系統。由於不同時期建置的資料具有不同的地籍坐標系統，因此常需要在兩地籍坐標系統間進行坐標轉換。目前，國內正積極將地籍資料由TWD67坐標系統轉換為TWD97坐標系統。而如何在TWD67與TWD97之間進行坐標轉換，整合不同地籍坐標系統間資料之聯繫與共享，一直是國內學者致力於研究的問題。在廣泛的討論當中，最常使用的方式為利用最小二乘法求解轉換參數。 近幾年來由於神經網路技術快速的發展，提供了我們在進行地籍坐標轉換研究時新的選擇。本研究目的在於嘗試利用神經網路方式進行TWD67與TWD97地籍坐標系統；同時為了提升神經網路的效用，及解決神經網路的黑盒子問題，本研究提出利用神經網路建構網格式地籍坐標轉換模式的方法。為了驗証本研究所提出之坐標轉換方法，利用三個大小不同的實驗區之共同點資料，由不同方式轉換所得的結果顯示，以純粹利用神經網路方式所得轉換結果為佳，而網格式地籍坐標轉換模式所得結果與利用最小二乘法求解結果不相上下。 / Currently, there are two cadastral coordinate systems used in Taiwan. They are TWD67 (Taiwan Datum 1967) and TWD97 (Taiwan Datum 1997) cadastral coordinate systems respectively. Frequently it is necessary to transform from one coordinate system to another. One of the most widely used method is Least-Squares with affine transformations. The artificial neural network (ANN) provides a new technology for cadastral coordinate transformation. The popularity of this methodology is rapidly growing. The greatest advantage of ANN is that it can be used very successfully with a huge quantity of data and free-model estimation that traditional transformation methods cannot be applied. In this research coordinate transformation between TWD67 and TWD97 with artificial neural network (ANN) and Least-Squares with affine transformations were examined. Besides, in order to overcome the so called ‘‘Black Box Problem’’ of ANN, algorithm of applying artificial neural network to develop regional grid-based cadastral coordinate transformation model was proposed. Three data sets with varied sizes from the Taiwan region are used to test the proposed algorithms. The test results show that the coordinate transformation accuracies using the ANN models are better than those of using other methods, such as, Least-Squares with affine transformations. The proposed algorithms and the detailed test results are presented in this paper.

對於神經網路系統將提出一個法則萃取的方式，並從神經網路中得到相關法則。在這裡我們所提到的方法是根據反函數的觀念而得到的。 / A rule-extraction method of the layered feed-forward neural networks is proposed here for identifying the rules suggested in the network. The method that we propose for the trained layered feed-forward neural network is based on the inversion of the functions computed by each layer of the network. The new rule-extraction method back-propagates regions from the output layer back to the input layer, and we hope that the method can be used further to deal with the predicament of ANN being a black box.

迴歸分析與類神經網路此兩種方法皆是預測領域上的主要工具。本論文嘗試在線性迴歸模式及非線性迴歸模式的條件下，隨機產生不同特性的資料以完整探討資料特性對迴歸分析與類神經網路之預測效果的影響。這些特性包括常態分配、偏態分配、不等變異、Michaelis-Menten關係模式及指數迴歸模式。 再者，我們使用區域搜尋法(local search methods)中的演化策略法(evolution strategies，ES)作為類神經網路的學習(learning)方法以提高其預測功能。我們稱這種類型的類神經網路為ESNN。 模擬結果顯示，ESNN確實可以取代常用來與迴歸分析做比較的倒傳遞類神經網路(back-propagation neural network，BPNN)，成為類神經網路的新選擇。針對不同特性的資料，我們建議：如果原始的資料適合以常態線性迴歸模式配適，則使用者可考慮使用迴歸方法做預測。如果原始的資料經由圖形分析或由檢定方法得知違反誤差項為均等變異之假設時，若能找到合適的權數，可使用加權最小平方法，但若權數難以決定時，則使用ESNN做預測。如果資料呈現韋伯偏態分佈時，可考慮使用ESNN或韋伯迴歸方法。資料適合以非線性迴歸模式做配適時，則選擇以ESNN做預測。 關鍵詞：迴歸分析，類神經網路，區域搜尋法，演化策略法類神經網路，倒傳遞類神經網路 / Both regression analysis and artificial neural networks are the main techniques for prediction. In this research, we tried to randomly generate different types of data, so as to completely explore the effect of data characteristics on the predictive performance of regression analysis and artificial neural networks. The data characteristics include normal distribution, skew distribution, unequal variances, Michaelis-Menten relationship model and exponential regression model. In addition, we used the evolution strategies, which is one of the local search methods for training artificial neural networks, to further improve its predictive performance. We name this type of artificial neural networks ESNN. Simulation studies indicate that ESNN could indeed replace BPNN to be the new choice of artificial neural networks. For different types of data, we commend that users can use regression analysis for their prediction if the original data is fit for linear regression model. When the residuals of the data are unequal variances, users can use weighted least squares if the optimal weights could be found. Otherwise, users can use ESNN. If the data is fit for weibull distribution, users can use ESNN or weibull regression. If the data is fit for nonlinear regression model, users can choose ESNN for the prediction. Keywords: Regression Analysis, Artificial Neural Networks, Local Search Methods, Evolution Strategies Neural Network (ESNN), Back-propagation Neural Network (BPNN)

本研究考慮重要且對台股大盤指數走勢有連動影響的因素，主要納入對台股有領頭作用的美國三大股市，那斯達克(NASDAQ)指數、道瓊工業(Dow Jones)指數、標準普爾500(S&P500)指數;其他對台股緊密連動效果的國際股票市場，香港恆生指數、上海證券綜合指數、深圳證券綜合指數、日經225指數;以及納入左右國際經濟表現的國際原油價格走勢，美國西德州原油、中東杜拜原油和歐洲北海布蘭特原油;在宏觀經濟因素方面則考量失業率、消費者物價指數、匯率、無風險利率、美國製造業重要指標的存貨/銷貨比率、影響貨幣數量甚鉅的M1B;在技術分析方面則納入多種重要的指標，心理線 (PSY) 指標、相對強弱(RSI) 指標、威廉(WMS%R) 指標、未成熟隨機(RSV) 指標、K-D隨機指標、移動平均線(MA)、乖離率(BIAS)、包寧傑%b和包寧傑帶狀寬度(BandWidth%);所有考量因素共計35項，因為納入重要因子比較多，所以完備性較高。 本研究先採用的贏者全拿(Winner-Take-All) 競爭學習策略的自組織映射網路(Self-Organizing Feature Maps, SOM)，藉由將相似資料歸屬到已身的神經元萃取出關聯分類且以計算距離來衡量神經元的離散特徵，對於探索大量且高維度的非線性複雜特徵俱有優良的因素相依性投射效果，將有利於提高預測模式精準度。在線性擬合部分則結合倒傳遞(Back-Propagation, BP)、Elman反饋式和徑向基底函數類網路(Radial-Basis-Function Network, RBF)模式為指數預測輸出，並對台股加權指數隔日收盤指數進行預測和評量。而在傳統的Elman反饋式網路只在隱藏層存在反饋機制，本研究則在輸入層和隱藏層皆建立反饋機制，將儲存在輸入層和隱藏層的過去時間資訊回饋給網路未來參考。在徑向基底函數網路方面，一般選取中心聚類點採用隨機選取方式，若能有效降低中心點個數，可降低網路複雜度，本研究導入垂直最小平方法以求取誤差最小的方式強化非監督式學習選取中心點的能力，以達到網路快速收斂，提昇網路學習品質。 研究資料為台股指數交易收盤價，日期自2001/1/2，至2011/10/31共2676筆資料。訓練資料自2001/1/2至2009/12/31，共2223筆;實證測試資料自2010/1/4至2011/10/31，計453個日數。主要評估指標採用平均相對誤差(AMRE)和平均絕對誤差 (AAE)。在考慮因子較多的狀況下，實證結果顯示，在先透過SOM進行因子聚類分析之後，預測因子被分成四個組別，分別再透過BP、Elman recurrent和RBF方法進行線性擬合，平均表現方面，以RBF模式下的四個群組因子表現最佳，其中RBF模式之下的群組4，其AMRE可達到0.63%，最差的AMRE則是群組1，約為1.05%;而Elman recurrent模式下的四組群組因子之ARME則介於1.01%和1.47%之間;其中預測效果表現最差則是BP模式的預測結果。顯示RBF具有絕佳的股價預測能力。最後，在未來研究建議可以運用本文獻所探討之其他數種類神經網路模式進行股價預測。 / In this study, we considering the impact factors for TWSE index tendency, mainly aimed at the three major American stock markets, NASDAQ index, Dow Jones index, S&P 500, which leading the Taiwan stock market trend; the other international stock markets, such as the Hong Kong Hang-Seng Index, Shanghai Stock Exchange Composite Index, Shenzhen Stock Exchange Composite Index, NIKKEI 225 index, which have close relationship with Taiwan stock market; we also adopt the international oil price trend, such as the West Texas Intermediate Crude Oil in American, the Dubai crude oil in Middle Eastern, North Sea Brent crude oil in European, which affects international economic performance widely; On the side of macroeconomic factors, we considering the Unemployed rate, Consumer Price Index, exchange rate, riskless rate, the Inventory to Sales ratio which it is important index of American manufacturing industry, and the M1b factor which did greatly affect to currency amounts; In the part of Technical Analysis index, we adopt several important indices, such as the Psychology Line Index (PSY), Relative Strength Index (RSI), the Wechsler Memory Scale—Revised Index (WMS%R), Row Stochastic Value Index (RSV), K-D Stochastics Index, Moving Average Line (MA), BIAS, Bollinger %b (%b), Bollinger Band Width (Band Width%);All factors total of 35 which we have considered the important factor is numerous, so the integrity is high. In this study, at first we adopt the Self-Organizing Feature Maps Network which based on the Winner-Take-All competition learning strategy, Similar information by the attribution to the body of the neuron has been extracted related categories and to calculate the distance to measure the discrete characteristics of neurons, it has excellent projection effect by exploring large and complex high-dimensional non-linear characteristics for all the dependency factors , would help to improve the accuracy of prediction models, would be able to help to improve the accuracy of prediction models. The part of the curve fitting combine with the back-propagation (Back-Propagation, BP), Elman recurrent model and radial basis function network (Radial-Basis-Function Network, RBF) model for the index prediction outputs, forecast and assessment the next close price of Taiwan stocks weighted index. In the traditional Elman recurrent network exists only one feedback mechanism in the hidden layer, in this study in the input and hidden layer feedback mechanisms are established, the previous information will be stored in the input and hidden layer and will be back to the network for future reference. In the radial basis function network, the general method is to selecting cluster center points by random selection, if we have the effectively way to reduce the number of the center points, which can reduces network complexity, in this study introduce the Orthogonal Least Squares method in order to obtain the smallest way to strengthen unsupervised learning center points selecting ability, in order to achieve convergence of the network fast, and improve network learning quality. Research data for the Trading close price of Taiwan Stock Index, the date since January 2, 2001 until September 30, 2011, total data number of 2656. since January 2, 2001 to December 31, 2009 a total number of 2223 trading close price as training data; empirical testing data, from January 4, 2010 to September 30, 2011, a total number of 433. The primary evaluation criteria adopt the Average Mean Relative Error (AMRE) and the Average Absolute Error (AAE). In the condition for consider more factors, the empirical results show that, by first through SOM for factor clustering analysis, the prediction factors were divided into four categories and then through BP, Elman recurrent and RBF methods for curve fitting, at the average performance , the four group factors of the RBF models get the best performance, the group 4 of the RBF model, the AMRE can reach 0.63%, the worst AMRE is group 1, about 1.05%; and the four groups of Elman recurrent model of ARME is between 1.01% and 1.47%; the worst prediction model is BP method. RBF has shown excellent predictive ability for stocks index. Finally, the proposal can be used in future studies of the literatures that we have explore several other methods of neural network model for stock trend forecasting.

面對市場未來趨勢的不確定性，投資者可以運用「投資組合保險」的概念，既能保障原本所投資的資產價值，又可以參與市場上漲時的獲利。本研究以類神經網路來研究證券市場的現象，一方面是已經有許多類神經網路在財務分析上的研究成果，另一方面是其具有學習以及預測的能力。 本研究首先探討投資組合保險策略，接著再比較投資組合保險策略在不同市況下的績效表現，隨後提出兩個階段的研究架構，經過設計與建置，以類神經網路模型進行對大盤未來漲跌型態的模擬預測，並利用預測的結果，輔助投資組合保險策略的決策，最後並將研究結果與大盤績效做綜合分析比較。 本研究的資料採取自台灣證券集中交易市場，期間為1991年1月3日至2002年12月31日，共3306個交易日，取大盤每日交易之歷史資料，經過處理後建立資料庫。類神經網路模型具有預測未來大盤漲跌區間的能力，在本研究所提出的漲跌區間劃分方式上，其預測正確率達到55%，預測的結果與實際漲跌完全相反的比例僅10%，其餘的35%為相鄰區間的預測誤差，其預測能力有助於投資組合保險策略的進行。 經過類神經網路模型輔助而進行的停損策略（SL），其年報酬率以及Sharpe Ratio，在大盤下跌的期間，兩個績效指標衡量結果皆為正值（21.125%＞0以及980.493＞0），充分發揮保險功能；而在大盤上漲的期間，兩個績效指標衡量結果皆優於大盤（46.544%＞17.137%以及393.808＞110.069）。 在年報酬率與Sharpe Ratio之間，本研究主張在探討投資組合保險時應著重風險的衡量，因此經過類神經網路模型輔助而進行的固定比例投資組合策略（CPPI），搭配槓桿乘數M值的調整，在大盤下跌的期間，其Sharpe Ratio依然可以維持正值，達到保險的效果，保護投資人的資產免於損失；而在大盤上漲的期間，其Sharpe Ratio更是高於大盤，可以享受資產價值提昇的獲利。 / Facing the uncertainty of the market trend, an investor can use the concept of “ Portfolio Insurance ” to protect the value of his portfolio in bear market and earn the benefit from bull market. There have been many researches about applying Neural Network in the financial analysis and Neural Network has the abilities to learn and forecast. This research evaluates the performances of the portfolio insurance strategies in different market trends. Then two-stage research structure has been designed and built. The first stage is forecasting the up-and-down trends of the equity market index by Neural network model. The second stage is using the forecasted results assisting the portfolio insurance decisions. Finally, the results of this research have been analyzed and compared with the benchmark. The Neural Network is able to forecast the future up-and-down trends. The accurate rate is 55%. During the bear market（2002）, the annual rate of return and Sharpe Ratio of the stop loss（SL） strategy which is assisted by NN are both positive（21.125%＞0 and 980.493＞0）. During the bull market（2001）, they both outperform the benchmark（46.544%＞17.137% and 393.808＞110.069）. The annual rate of return is more important than Sharpe Ratio because the risk measurement is an important factor in portfolio insurance strategy. Sharpe Ratios of the CPPI strategy which is assisted by NN outperform the benchmark in both above mentioned bear and bull market. In short, the SL and CPPI strategy assisted by NN not only protect the value of the portfolio from losing in bear market but also gain profit in bull market, so they are the ideal portfolio insurance strategies.

當市場上有訊息產生時，由於市場機制的限制或其他因素影響，例如投資人的心理，導致不同股票在傳導訊息時存在時間上的落差，因此產生反應訊息領先其他股票的指標股，在本研究中將利用兩種方法：矩陣自我迴歸及類神經網路檢測市場上是否存有指標股、其特性為何？及指標股是否具有穩定性。 　 實證結果發現規模較大及流動性較高的股票在市場呈現多頭行情時其訊息傳遞確實具有領先效果且領先時期達六至十二個營業日，而券商於報章雜誌上推薦之指標股並不具備領先反應訊息的能力。此外，在市場呈現空頭時期，所有股票反應訊息的速度幾乎一致。

This research employs five predicting variables to implementing the market timing strategy. These five variables are E/P1, E/P2, B/M, CP and GM. The investment performances of market timing under a variety of investment horizons are examined. There are four different forecasting horizons, which are one-month, three-month, six-month, and twelve-month investment horizons. Both the linear approach and artificial neural networks are employed to forecasting the market. The artificial neural network is employed with a view to capture the non-linearity property embedded in the market. The results are summarized as follows. (1) Both the linearity and nonlinear approaches are able to outperform the market. According to the results of Cumby-Modest test, they do have the market timing ability. (2) In the simple regression models, the performance of CP is relatively well compared to those of other variables. (3) The correct prediction rate increases as the investment horizon increases. (4) The performance of the expanding window approach is on average inferior to that of the moving window approach. (5) In the simulations of timing abilities over the period of May, 1991 to December, 1997. The multiple regression models has the best performance for the cases of one-month, three-month, and six-month investment horizons. On the other hand, BP(1) has the best performance for the case of one-year investment horizon. Contents Chapter 1 Introduction ……………………………………… 1 1.1 Background……………………………………………………………. 1 1.2 Motivations and objectives…………………………………………….3 1.3 Thesis organization ………………………………………………….. 4 Chapter 2 Literature Review…………………………………6 2.1 Previous studies on market timing……………………………………. 6 2.2 Predicting variables…………………………………………………… 8 2.3 Artificial Neural Networks……………………………………………10 2.4 Back Propagation Neural Networks…………………………………..11 2.5 Applications of ANNs to financial fields………………….………….12 Chapter 3 Data and Methodology……………………….….15 3.1 Data………………………………………………………………..….15 3.2 Linear approaches to implementing market timing strategy……….…18 3.3 ANNs to implementing market timing strategy…………..…………..23 Chapter 4 Results on Timing Performance……………..…26 4.1 Performance of linear approach………………………………………26 4.2 Performance of ANNs………………………………………………...38 4.3 Performance evaluation……………………………………………….39 Chapter 5 Summary…………………………………………54 5.1 Conclusions……………………………………………………….….54 5.2 Future works…………………………………………………………55 Appendix……………………………………………………..56 References……………………………………………………57 / This research employs five predicting variables to implementing the market timing strategy. These five variables are E/P1, E/P2, B/M, CP and GM. The investment performances of market timing under a variety of investment horizons are examined. There are four different forecasting horizons, which are one-month, three-month, six-month, and twelve-month investment horizons. Both the linear approach and artificial neural networks are employed to forecasting the market. The artificial neural network is employed with a view to capture the non-linearity property embedded in the market. The results are summarized as follows. (1) Both the linearity and nonlinear approaches are able to outperform the market. According to the results of Cumby-Modest test, they do have the market timing ability. (2) In the simple regression models, the performance of CP is relatively well compared to those of other variables. (3) The correct prediction rate increases as the investment horizon increases. (4) The performance of the expanding window approach is on average inferior to that of the moving window approach. (5) In the simulations of timing abilities over the period of May, 1991 to December, 1997. The multiple regression models has the best performance for the cases of one-month, three-month, and six-month investment horizons. On the other hand, BP(1) has the best performance for the case of one-year investment horizon.