[논문]주식 시장 예측을 위한 π-퍼지 논리와 SVM의 최적 결합

다오두안훙; 안현철

doi:10.13088/jiis.2014.20.4.43

[국내논문] 주식 시장 예측을 위한 π-퍼지 논리와 SVM의 최적 결합
An Optimized Combination of π-fuzzy Logic and Support Vector Machine for Stock Market Prediction 원문보기

지능정보연구 = Journal of intelligence and information systems, v.20 no.4, 2014년, pp.43 - 58

다오두안훙 (국민대학교 비즈니스IT전문대학원) , 안현철 (국민대학교 비즈니스IT전문대학원)

초록
AI-Helper

최근 정보기술의 발전으로 복잡하고 방대한 양의 주가 데이터에 대한 실시간 분석이 가능해지면서 인공지능 기법을 활용해 주식 시장의 등락을 예측하고, 이를 기반으로 매매 거래를 수행하는 트레이딩 시스템에 대한 세간의 관심이 높아지고 있다. 본 연구는 이러한 트레이딩 시스템의 시장 예측 알고리즘으로 활용될 수 있는 새로운 주식 시장 등락 예측 모형을 제시한다. 본 연구의 제안 모형은 ${\pi}$-퍼지 논리를 이용해 모든 입력변수의 차원을 low, medium, high로 퍼지변환한 입력값을 대상으로 Support Vector Machine(SVM)을 적용하여 익일 시장의 등락을 예측하도록 설계되었다. 그런데 이 경우 입력변수의 수가 3배로 늘어나기 때문에, 적절한 입력변수의 선택이 요구된다. 이에 본 연구에서는 유전자 알고리즘을 활용하여 입력변수 선택 집합을 최적화하도록 하였으며, 동시에 ${\pi}$-퍼지 논리 및 SVM에 적용되는 조절 파라미터들의 값도 함께 최적화 하도록 하였다. 모형의 성능을 검증하기 위해, 본 연구에서는 지난 2004년부터 2013년까지의 10년치 국내 주식시장 데이터를 기반으로 한 KOSPI 200 지수의 등락 예측에 제안모형을 적용해 보았다. 이 때, 비교모형으로 로지스틱 회귀모형, 다중판별분석, 의사결정나무, 인공신경망, SVM, 퍼지SVM 등도 함께 적용시켜 성과를 정밀하게 검증해 보고자 하였다. 그 결과, 제안모형이 예측 정확도는 물론 투자수익률(Return on Investment) 측면에서도 다른 모든 비교모형들에 비해 월등히 우수한 성능을 보임을 확인할 수 있었다.

Abstract ▼ AI-Helper

As the use of trading systems has increased rapidly, many researchers have become interested in developing effective stock market prediction models using artificial intelligence techniques. Stock market prediction involves multifaceted interactions between market-controlling factors and unknown random processes. A successful stock prediction model achieves the most accurate result from minimum input data with the least complex model. In this research, we develop a combination model of ${\pi}$-fuzzy logic and support vector machine (SVM) models, using a genetic algorithm to optimize the parameters of the SVM and ${\pi}$-fuzzy functions, as well as feature subset selection to improve the performance of stock market prediction. To evaluate the performance of our proposed model, we compare the performance of our model to other comparative models, including the logistic regression, multiple discriminant analysis, classification and regression tree, artificial neural network, SVM, and fuzzy SVM models, with the same data. The results show that our model outperforms all other comparative models in prediction accuracy as well as return on investment.

Keyword

AI 본문요약
AI-Helper

* AI 자동 식별 결과로 적합하지 않은 문장이 있을 수 있으니, 이용에 유의하시기 바랍니다.

제안 방법

This study proposes a binary classification model that combines the π-fuzzy logic and support vector machine (SVM) models for stock market prediction.
To enhance the prediction quality of the model, a genetic algorithm (GA) is used to find the optimized values of the parameters and to optimize the feature selection. The performance of the proposed model is compared to comparative models such as the logistic regression (LOGIT), multiple discriminant analysis (MDA), classification and regression tree (CART), artificial neural network (ANN), SVM, and fuzzy SVM models.
To evaluate the performance of the proposed model, we compare the performances of our model to other comparative models using LOGIT, MDA, CART, ANN, SVM, and fuzzy SVM on the same data. LOGIT, MDA, and CART are tested using IBM SPSS Statistics 20, and ANN using Neuroshell2.
Although it is important to accurately predict the directions of the stock market, it is more important to yield better ROI using the prediction model in the trading systems domain. For this reason, we apply our model and other comparative models to the hold-out dataset, and simulate trading transactions according to the signals from these models in order to measure their ROIs. Table 6 depicts the ROIs of each model.
Third, the general applicability of the proposed model should be tested further. Although we applied our model to stock market prediction in this study, it can be applied to any domain that requires accurate prediction.

대상 데이터

The data used in this study consist of 2,210 daily observations of the KOSPI 200. It covers a 10-year period, from January 2, 2004, to December 30, 2013.
The data were divided into two subsets: training and hold-out datasets. The data from 2004 to 2011 (1,778 samples, about 80%) were used as the training dataset, and the data from the remaining two, more recent, years (493 samples, about 20%) used as the hold-out dataset.
The data were divided into two subsets: training and hold-out datasets. The data from 2004 to 2011 (1,778 samples, about 80%) were used as the training dataset, and the data from the remaining two, more recent, years (493 samples, about 20%) used as the hold-out dataset. Table 2 shows the number of cases for the training and hold-out datasets in each year.
Our proposed model optimizes the feature selection, kernel, and π-fuzzy parameters simultaneously. In order to validate the usefulness of our model, we applied it to a Korean stock market dataset covering 10 years. As a result, we found that our proposed model showed higher prediction accuracy and ROI than other conventional models such as LOGIT, MDA, CART, ANN, SVM, and fuzzy SVM.

이론/모형

The experimental system was developed using LIBSVM v2.8 (Chang and Lin, 2011), Evolver v5.5, and Microsoft Visual Basic for Applications (VBA). Evolver, a commercial software application, was used for implementing GA, and LIBSVM used for training SVM classifiers.

성능/효과

(2009) proposed an integrated system, CBDWNN, by combining dynamic time windows, case-based reasoning (CBR), and neural networks (NN) for stock trading prediction. The empirical results show that CBDWNN overtakes models that use CBR or BPN alone. An improved bacterial chemotaxis optimization (IBCO), integrated into the backpropagation (BP) artificial neural network forecasting model by Yudong and Lenan (2009) showed better performance than other methods in terms of learning ability and generalization.
Kim and Ahn (2011) used GA to optimize the instance selection process with simultaneous parameter optimization. The ISVM model (SVM with instance selection) was compared to several other comparative models, including LOGIT, backpropagation neural networks (BPN), nearest neighbor (1-NN), conventional SVM, and SVM with optimized parameters (PSVM), to prove its outstanding performance. An intelligent trading system created by Kim and Ahn (2010) was designed to use both technical indicators and other non-price variables of the market.
As shown in Table 4, our proposed model outperforms all the others. In detail, it achieves prediction accuracy higher than LOGIT, MDA, CART, ANN, SVM, and fuzzy SVM by 9.13%, 10.14%, 10.95%, 4.26%, 2.84%, and 2.63%, respectively, for the hold-out dataset.
45% per year on average. The yearly ROI of our model is more than 10% higher than the second-best (fuzzy SVM, 19.21%), and more than 24% higher than the benchmark strategy (5.35%).
In order to validate the usefulness of our model, we applied it to a Korean stock market dataset covering 10 years. As a result, we found that our proposed model showed higher prediction accuracy and ROI than other conventional models such as LOGIT, MDA, CART, ANN, SVM, and fuzzy SVM. In particular, the ROI of our model was found to be more than five times higher than the benchmark strategy.

참고문헌 (40)

Ahn, H. and H. Y. Lee, "A combination model of Multiple Artificial Intelligence techniques based on Genetic Algorithms for Investment Decision Support Aid: An Application to KOSPI", The e-Business Studies, Vol.10, No.1 (2009), 215-236.
Atsalakis, G. and K. Valavanis, "Neuro-fuzzy and technical analysis for stock prediction," Working paper, 2006.
Atsalakis, G. S. and K. P. Valavanis, "Surveying stock market forecasting techniques - Part II: Soft computing methods", Expert Systems with Applications, Vol.36, No.3(2009), 5932-5941.

상세보기
Chang, C.-C. and C.-J. Lin, "LIBSVM: a library for support vector machines", ACM Transactions on Intelligent Systems and Technology, Vol.2, No.3(2011), 27:1-27:27. Software available at http://www.csie.ntu.edu.tw/-cjlin/libsvm.
Chang, P.-C., C.-H. Liu, J.-L. Lin, C.-Y. Fan, and C. S. P. Ng, "A neural network with a case based dynamic window for stock trading prediction", Expert Systems with Applications, Vol.36, No.3(2009), 6889-6898.

상세보기
Chen, C.-H., T.-K. Liu, J.-H. Chou, C.-H. Tasi, and H. Wang, "Optimization of teacher volunteer transferring problems using greedy genetic algorithms", Expert Systems with Applications, Vol.42, No.1(2015), 668-678.

상세보기
Constantinou, E., R. Georgiades, A. Kazandjian, and G. P. Kouretas, "Regime switching and artificial neural network forecasting of the Cyprus Stock Exchange daily returns", International Journal of Finance and Economics, Vol.11, No.4(2006), 371-383.

상세보기
Cortes, C. and V. Vapnik, "Support-vector networks", Machine Learning, Vol.20, No.3(1995), 273-297

상세보기
Debashish D. and S. U. Mohammad, "Data mining and Neural network techniques in Stock market prediction: A methodological review", International Journal of Artificial Intelligence & Applications (IJAIA), Vol.4, No.1(2013), 117-127.
Esen, I. and M. A. Koc, "Optimization of a passive vibration absorber for a barrel using the genetic algorithm", Expert Systems with Applications, Vol.42, No.2(2015), 894-905.

상세보기
Goldberg, D., Genetic algorithms in search, optimization, and machine learning, Addison-Wesley, 1989.
Gordini, N., "A genetic algorithm approach for SMEs bankruptcy prediction: Empirical evidence from Italy", Expert Systems with Applications, Vol.41, No.14(2014), 6433-6445.

상세보기
Harnett, D. L. and A. K. Soni, Statistical methods for business and economics, Addison-Wesley. Massachusetts, MA, (1991).
Holland, J. H., "Genetic Algorithms", Scientific American, Vol.267, No.1(1992), 66-72.

상세보기
Huang, H.-X., J.-C. Li, and C.-L. Xiao, "A proposed iteration optimization approach integrating backpropagation neural network with genetic algorithm", Expert Systems with Applications, Vol.42, No.1(2015), 146-155.

상세보기
Jhee, W. C. and J. K. Lee, "Performance of Neural networks in managerial forecasting", Intelligent Systems in Accounting, Finance, and Management, Vol.2, No.1(1993), 55-71.

상세보기
Jo, H., "The integrated methodology of artificial intelligence and statistical methods for bankruptcy prediction," Ph.D. Dissertation, Dept. of Management Engineering, KAIST, 1999.
Kanas, A. and A. Yannopoulos, "Comparing linear and nonlinear forecasts for stock returns", International Review of Economics and Finance, Vol.10, No.4(2001), 383-398.

상세보기
Kim, K.-j. and H. Ahn, "Optimization of Support Vector Machines for Financial Forecasting", Journal of Intelligence and Information Systems, Vol.17, No.4(2011), 241-254.
Kim, K.-j., and H. Ahn, "Simultaneous optimization of artificial neural networks for financial forecasting", Applied Intelligence, Vol.36, No.4(2012), 887-898.

상세보기
Kim, S.-W. and H. Ahn, "Development of an Intelligent Trading System Using Support Vector Machines and Genetic Algorithms", Journal of Intelligence and Information Systems, Vol.16, No.1(2010), 71-92.
Koulouriotis, D. E., I. E. Diakoulakis, D. M. Emiris, and C. D. Zopounidis, "Development of dynamic cognitive networks as complex systems approximators: Validation in financial time series", Applied Soft Computing, Vol.5, No.2(2005), 157-179.

상세보기
Lee, J. K., "Integration and competition of AI with quantitative methods for decision support", Expert Systems with Applications, Vol.1, No.4(1990), 329-333.

상세보기
Lendasse, A., E. de Bodt, V. Wertz, and M. Verleysen, "Non-linear financial time series forecasting - Application to the Belgium 20 Stock Market Index", European Journal of Economical and Social Systems, Vol.14, No.1(2000), 81-91.

상세보기
Liang, T.-P., J. S. Chandler, and I. Han, "Integrating statistical and inductive learning methods for knowledge acquisition", Expert Systems with Applications, Vol.1, No.4(1990), 391-401.

상세보기
Majhi, R., G. Panda, B. Majhi, and G. Sahoo, "Efficient prediction of stock market indices using adaptive bacterial foraging optimization (ABFO) and BFO based techniques", Expert Systems with Applications, Vol.36, No.6 (2009), 10097-10104.

상세보기
Marczyk, A., Genetic algorithms and evolutionary computation, The TalkOrigins Archive, 2004. Available at http://www.talkorigins.org/faqs/genalg/genalg.html (Downloaded 1 November, 2014).
Medsker, L. and E. Turban, "Integrating expert systems and neural computing for decision support", Proceedings of the Twenty-Seventh Hawaii International Conference on System Sciences, (1994), 656-665.
Mitchell, M., An Introduction to Genetic Algorithms, MIT Press, Massachusetts, 1998.
Mohammadian, M. and M. Kingham, "An adaptive hierarchical Fuzzy logic system for modelling of Financial Systems", Intelligent Systems in Accounting, Finance and Management, Vol.12, No.1(2004), 61-82.

상세보기
Novak, V., I. Perfilieva, and J. Mockor, Mathematical Principles of Fuzzy Logic, Kluwer Academic, Dordrecht, 1999.
Pal, S. K. and P. K. Pramanik, "Fuzzy measures in determining seed points in clustering", Pattern Recognition Letters, Vol.4, No.3(1986), 159-164.

상세보기
Fernandez-Rodriiguez F., C. Gonzalez-Martel, and S. Sosvilla-Rivebo, "On the profitability of technical trading rules based on artificial neural networks: Evidence from the Madrid Stock Market", Economics Letters, Vol.69, No.1(2000), 89-94.

상세보기
Perez-Rodriguez, J. V., S. Torra, and J. Andrada-Felix, "STAR and ANN models: Forecasting performance on the Spanish "Ibex-35" stock index", Journal of Empirical Finance, Vol.12, No.3(2004), 490-509.
Steiner, M., and H.-G. Wittkemper, "Portfolio optimization with a neural network implementation of the coherent market hypothesis", European Journal of Operational Research, Vol.100, No.1(1997), 27-40.

상세보기
Tan, T. Z., C. Quek, and G. S. Ng, "Brain inspired genetic complimentary learning for stock market prediction", Proceedings of the IEEE congress on evolutionary computation, Vol.3(2005), 2653-2660.
Zhang, Y. and L. Wu, "Stock market prediction of S&P 500 via combination of improved BCO approach and BP neural network", Expert Systems with Applications, Vol.36, No.5(2009), 8849-8854.

상세보기
Yumlu, M. S., F. S. Gurgen, and N. Okay, "Turkish stock market analysis using mixture of experts", Proceedings of Engineering of Intelligent Systems (EIS), (2004).
Yumlu, S., F. S. Gurgen, and N. Okay, "A Comparison of global, recurrent and smoothed-piecewise neural models for Istanbul Stock Exchange prediction", Pattern Recognition Letters, Vol.26, No.13 (2005), 2093-2103.

상세보기
Zadeh, L.A., "Fuzzy sets", Information and Control, Vol.8, No.3(1965), 338-353.

상세보기

저자의 다른 논문 :

LOADING...

표제어: PCR

동의어: Packet Collision Rate

용어 설명 출처 목록 (6)

용어 설명: PCR은 세균 특이성이 있는 primer를 이용하여 적은 수의 세균이 있을지라도 쉽게 검출할 수 있는 유용한 방법이며, 이를 이용하여 구강 내 치면세균막이나 타액에서 직접 세균을 검출할 수 있게 되었다[8].

내보내기 구분	파일저장 인쇄 메일전송
구성항목	기본정보 상세정보 관리번호, 논문명, 저널/프로시딩명, 저자 , 발행년, 권, 호, 시작페이지, 끝페이지, 발행기관 관리번호, 논문명, 대등논문명, 저자 , 저널/프로시딩명, 발행기관, 발행년, 발행언어, 권, 호, 시작페이지, 끝페이지, ISBN, ISSN, 주제분야, 키워드, 초록(한글), 초록(영문), 저자(소속기관)
저장형식	Text(ASCII format) Excel format RefWorks Direct Export RIS format (for Reference Manager, ProCite, EndNote), Scholar's Aids, Mendeley
메일정보	받는사람 (필수) @ 보내는사람 (선택) @ 제목 내용 KISTI 검색결과 이메일 서비스
안내	총 건의 자료가 검색되었습니다. 다운받으실 자료의 인덱스를 입력하세요. (1-10,000) 검색결과의 순서대로 최대 10,000건 까지 다운로드가 가능합니다. 데이타가 많을 경우 속도가 느려질 수 있습니다.(최대 2~3분 소요) 다운로드 파일은 UTF-8 형태로 저장됩니다. 파일의 내용이 제대로 보이지 않을실 때는 웹브라우저 상단의 보기 -> 인코딩 -> 자동선택 여부를 확인하십시오. ~ Text(ASCII format) Excel format

연합인증