EXAMINING THE EFFECT OF BIAS ON STOCK PRICE PREDICTION PERFORMANCE

Yıl 2017, , 97 - 113, 29.12.2017

Mehmet Özçalıcı

https://doi.org/10.17065/huniibf.372406

Öz

Artificial Neural Network
models are successfully applied in the field of stock price predicting. Bias
values are important factor that can affect the performance of neural networks.
Although its importance, the effect of bias on stock price forecasting has not
been investigated in literature. The purpose of this study is to examine the
effect of bias on stock price predicting performance. For
this purpose, historical price and volume information of stocks listed in
BIST100 Index is used. By means of these data, 201 technical indicators are
calculated. Activation function type in forecasting model and number of neurons
in the hidden layer and variable selection are optimized with the Harmony Search
Algorithm, a population-based meta-heuristic optimization method. RMSE and hit
rate measurements are used as performance indicators. As a result, no
statistically significant performance difference was found between biased and
un-biased neural network models. 
However, it has been determined that the training of the models for
un-biased models has been completed in a shorter time.  

Anahtar Kelimeler

Artificial neural networks, stock price predicting, harmony search, bias value

EŞİK DEĞERİNİN HİSSE SENEDİ FİYAT TAHMİN PERFORMANSINA ETKİSİNİN İNCELENMESİ

Öz

Yapay
sinir ağları hisse senedi fiyat tahmininde başarı ile uygulanmaktadır. Eşik
değerleri yapay sinir ağlarının performansını etkileyebilecek önemli bir
unsurdur. Bununla beraber yapay sinir ağları modelinde yer alan eşik değerinin
tahmin performansı üzerindeki etkisinin incelendiği bir çalışmaya literatürde
rastlanma-mıştır. Bu çalışmanın amacı eşik değerinin hisse senedi fiyat tahmin
performansı üzerindeki etkisini incelemektir. Bu amaçla, BIST100 endeksinde
listelenen 100 adet hisse senedine ilişkin 1 Ocak 2014 ve 30 Haziran 2015
tarihleri arasındaki tarihi fiyat ve işlem hacmi bilgileri kullanılmıştır. Bu
verilerden yola çıkmak suretiyle 201 adet teknik gösterge hesaplanmıştır. Yapay
sinir ağlarındaki aktivasyon fonksiyonu çeşidi, ara katmanda yer alması gereken
nöron sayısı ve değişken seçimi, popülasyon tabanlı meta-sezgisel bir yöntem
olan Harmoni Arama algoritması ile optimize edilmiştir. Performans ölçüsü
olarak RMSE ve doğru tahmin oranı değerleri kullanılmıştır. Sonuçta eşik
değerinin yer aldığı modeller ile eşik değerinin yer almadığı modeller arasında
istatistiksel açıdan anlamlı bir performans farklılığı tespit edilememiştir.
Bununla beraber eşik değerinin yer almadığı modellerin eğitiminin daha kısa
sürede tamamlandığı belirlenmiştir. 

Anahtar Kelimeler

Yapay sinir ağları, hisse senedi fiyat tahmini, harmoni arama, eşik değeri

Kaynakça

• Achelis, S.B. (2001), Technical Analysis From A to Z, New York: McGrawHill.
• Adhikari, R., R.K. Agrawal (2014), "A Combination of Artificial Neural Network and Random Walk Models for Financial Time Series Forecasting", Neural Computing and Applications, 24(6), 1441–1449.
• Asadi, S., E. Hadavandi, F. Mehmanpazir, M.M. Nakhostin (2012), "Hybridization of Evolutionary Levenberg-Marquardt Neural Networks and Data Pre-Processing for Stock Market Prediction", Knowledge-Based Systems, 35, 245–258.
• Atsalakis, G.S., E.E. Protopapadakis, K.P. Valavanis (2015), "Stock Trend Forecasting in Turbulent Market Periods Using Neuro-Fuzzy Systems", Operational Research, 16(2), 245-269.
• Atsalakis, G.S., K.P. Valavanis (2009a), "Forecasting Stock Market Short-Term Trends Using a Neuro-Fuzzy Based Methodology", Expert Systems with Applications, 36(7), 10696–10707.
• Atsalakis, G.S., K.P. Valavanis (2009b), "Surveying Stock Market Forecasting Techniques-Part II: Soft Computing Methods", Expert Systems with Applications, 36(3 PART 2), 5932–5941.
• Aygören, H., H. Sarıtaş, T. Moralı (2012), "İMKB 100 Endeksinin Yapay Sinir Ağları ve Newton Nümerik Arama Modelleri ile Tahmini", Uluslararası Alanya İşletme Fakültesi Dergisi, 4(1), 73–88.
• Dai, W., Y.E. Shao, C.J. Lu (2012), "Incorporating Feature Selection Method into Support Vector Regression for Stock Index Forecasting", Neural Computing and Applications, 1551–1561.
• Fama, E.F. (1965), "The Behavior of Stock-Market Prices", Journal of Business, 38(1), 34-105.
• Hagan, M.T., H.B. Demuth, M.H. Beale (1996), Neural network design. Boston Massachusetts PWS (Vol. 2), Retrieved from http://ecee.colorado.edu/academics/schedules/ECEN5120.pdf
• Hsu, C.M. (2013), "A Hybrid Procedure with Feature Selection for Resolving Stock/Futures Price Forecasting Problems", Neural Computing and Applications, 22(3-4), 651–671.
• Kara, Y., M.A. Boyacioglu, Ö.K. Baykan (2011), "Predicting Direction of Stock Price Index Movement Using Artificial Neural Networks and Support Vector Machines: The Sample of the Istanbul Stock Exchange", Expert Systems with Applications, 38(5), 5311–5319.
• Kaufman, P.C. (1998), Trading Systems and Methods (3rd Edition), Canada: Wiley & Sons.
• Mahmud, M.S., P. Meesad (2015), "An Innovative Recurrent Error-Based Neuro-Fuzzy System with Momentum for Stock Price Prediction", Soft Computing, 20(10), 4173-4191.
• Özçalıcı, M. (2015), Hisse Senedi Fiyat Tahminlerinde Bilgi İşlemsel Zeka Yöntemleri: Uzman Bir Sistem Aracılığıyla BIST Uygulaması, Kahramanmaraş: Kahramanmaraş Sütçü İmam Üniversitesi, Doktora Tezi.
• Özdemir, Ö. (2015), "An Improved Method for Forecasting Borsa Istanbul 100 Index by Neural Network Based Fuzzy Time Series and Particle Swarm Optimization", In 16th International Symposium on Econometrics, Operations Research and Statistics, Edirne.
• Palit, A. K., D. Popovic (2005), Computational Intelligence in Time Series Forecasing: Theory and Engineering Applications, London: Springer-Verlag.
• Sarıkaya, G. (2014), "Forecasting Bist National-100 Index by Using Artificial Neural Network and Regression Models", In 15th International Symposium on Econometrics, Operations Research and Statistics, 155–166, Isparta.
• Wu, Y., R.M. Rangayyan (2007), "An Algorithm for Evaluating the Performance of Adaptive Filters for the Removal of Artifacts in ECG Signals", In 2007 Canadian Conference on Electrical and Computer Engineering, 864–867, IEEE.
• Yakut, E., B. Elmas, S. Yavuz (2014), "Yapay Sinir Ağları ve Destek Vektör Makineleri Yöntemleriyle Borsa Endeksi Tahmini", Süleyman Demirel Üniversitesi, İktisadi ve İdari Bilimler Fakültesi Dergisi, 19(1), 139–157.
• Yang, J., W. Wu (2007), "Is Bias Dispensable for Fuzzy Neural Networks?", Fuzzy Sets and Systems, 158(24), 2757–2762.
• Yang, X. S. (2009), "Harmony Search as a Metaheuristic Algorithm", Studies in Computational Intelligence, 191, 1–14.
• Yu, L., S. Wang, K.K. Lai (2005), "Mining Stock Market Tendency Using GA-Based Support Vector Machines", Lecture Notes in Computer Science (including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 3828 LNCS, 336–345.
• Zhai, F., Q. Wen, Z. Yang Y. Song (2010), "Hybrid Forecasting Model Research on Stock Data Mining", New Trends in Information Science and Service Science (NISS), 2010 4th International Conference on, Gyeongju, 630-633.