整合電腦及通訊網路技術之分散式系統(Digtributed System)，係將分散各地的電腦 系統及相關的資訊設備透過通訊網路連接在一起，以形成能夠相互效換資訊，分享資 源，並具備分散式處理能力的系統。由於電腦及通訊網路技術突破性的進展，在許多 情況下，已使分散式系統較傳統的集中式系統具有低成本、高可靠度、容易擴充、效 能較佳等優勢。 本研究根據分散式系統的基本架構，使用排隊網路(Queueing Network)分析技術，建 立一套一般性的分散式系統模式，此模式可用以評估分散式系統運作之效能(Perfor- mance)，探討影響分散式系統效能的因素。同時，本研究進一步利用系統模擬(Syst- em Simulation)技術分析分散式系統模式，並進行分析結果的比較與驗證，以做為規 劃與設計分散式系統的參考依據。為說明其應用，我們並以校園線上即時選課資訊網 路系統為例，找出影響其效能的主要因素及可接受的設計方案，以做為校園資訊網路 整體系統架構設計決策的參考依據。 本研究之預期貢獻主要在於提供一套簡單經濟且具有相當程度準確性的方法，藉以模 擬、預測或評估分散式系統之行為與效能，這套方法可以廣泛應用在各種不同組態下 的分散式系統。

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.

本研究設計了一些實驗來檢測學習完混沌資料的神經網路系統是否為混沌系統，驗證的方法是檢驗是否具有混沌資料的四個特性，這四個特性包括：有限性、非週期性、確定性、及對初始條件的敏感依賴。同時，更進一步地利用上述學習完的網路系統來預測所學習的混沌模型，這麼做的目的是想要了解：學習後的網路系統是一個混沌系統時，與學習後網路系統不是一個混沌系統時，其預測能力的比較。 此外，我們亦從理論上證明：學習完混沌資料後的神經網路系統無法重建其所學習的混沌模型。然而，有時網路系統卻能夠模擬成一個混沌系統；如果使用模擬成混沌系統的神經網路來預測具有混沌現象的資料，換句話說，也就可能是使用一個混沌系統去預測另一個混沌系統，根據混沌的特性 -- 對初始條件的敏感依賴，這樣的預測應該會造成相當大的誤差；不過，從本研究的實驗中發現，無論學習後的神經網路系統是否為一個混沌系統，對其預測能力並無顯著的影響。 本論文希望能給「用神經網路系統來預測具有混沌現象的金融市場或其他領域」一些貢獻與幫助。 / This paper uses some experimental designs to detect if the Neural Networks system after learning the chaotic data is a chaotic system. That is verified via testing four characteristics in chaotic data, inclusive of boundedness, determinism, aperiodicity and sensitive dependence on initial conditions. Further, this paper uses the result above to predict the learned chaotic model. The purpose is to probe into if the Neural Networks system after learning the chaotic data is a chaotic system and is used to predict, how good the short-term and the long-term predictions will be? And, compare with if the Neural Networks system after learning the chaotic data is not a chaotic system and is used to predict, how large the error will be? We present the Neural Network systems after learning the chaotic data never can rebuild the learned chaotic model. But, sometimes the Neural Network system would mimic as a chaotic system. So, if we take Neural Network system to predict something with chaotic phenomena, it is possible to use one chaotic system to predict another chaotic system. According to the property of sensitive dependence on initial conditions, it should make large errors. However, from the experiments we design, we find whether the Neural Network system after learning is a chaotic system or not, it has no influence on its predicting effect. This hint is applied to use ANN to predict in financial markets or other areas with chaotic phenomenon.