Finansal Karar Süreçlerinde Makine Öğrenimi Tabanlı Tahmin ve Strateji Geliştirme Yaklaşımları

Yıl 2025, Cilt: 7 Sayı: 2, 213 - 237, 30.08.2025

Mahmut Emin Çevik , Bengü Vuran

https://doi.org/10.38009/ekimad.1710303

Öz

Finansal piyasaların doğasında bulunan belirsizlik nedeniyle, yalnızca sezgilere dayalı yatırım kararları önemli riskler taşımaktadır. Bu nedenle, bu çalışma, makine öğrenimi tabanlı tahmin tekniklerinin yatırım karar süreçlerine nasıl entegre edilebileceğini araştırmayı amaçlamaktadır. Random Forest, XGBoost ve LSTM modellerini içeren karşılaştırmalı analizimizde, Random Forest modeli, yaklaşık 24 endeks puanlık ortalama mutlak hata (MAE) ile daha düşük tahmin hatası ve daha tutarlı sinyallerle öne çıkmıştır. Bu sinyaller yatırım stratejilerine dönüştürüldüğünde, yalnızca tahmin doğruluğunda değil, aynı zamanda %95,51 toplam getiri ve 8,06 Sharpe oranı gibi risk-ayarlı getirilerde de anlamlı iyileşmeler gözlenmiştir. Ancak, bu bulguların çalışmada kullanılan veri seti ve modelleme seçimlerine özgü olduğu ve farklı piyasa koşullarında veya veri rejimlerinde sonuçların değişebileceği vurgulanmalıdır.

Anahtar Kelimeler

Finansal zaman serisi , Makine öğrenimi , Random Forest , XGBoost , LSTM , Teknik analiz , Alım-satım stratejisi , Geriye dönük test , Sharpe oranı

Forecasting and Strategy Design in Financial Decision-Making Using Machine Learning

Öz

Owing to the inherent uncertainty of financial markets, investment decisions based solely on intuition involve substantial risk. Accordingly, this study investigates the integration of machine learning-based forecasting techniques into investment decision-making processes. In the comparative analysis conducted with Random Forest, XGBoost, and LSTM models, the Random Forest model demonstrated superior performance, yielding a lower mean absolute error (MAE) of approximately 24 index points and generating more consistent predictive signals. When these signals were translated into trading strategies, the model exhibited substantial improvements not only in predictive accuracy but also in risk-adjusted performance, achieving a total return of 95.51% and a Sharpe ratio of 8.06. Nevertheless, it should be noted that these findings are contingent upon the specific dataset and modeling configurations employed, and may vary across different market environments and data regimes.

Anahtar Kelimeler

Financial time series , Machine learning , Random Forest , XGBoost , LSTM , Technical analysis , Trading strategy , Backtesting , Sharpe ratio

Kaynakça

• Abdullah, M., Ahmed, F., Ve Khan, R. (2022), Machine Learning Approaches For Financial Market Prediction: A Review, Journal Of Intelligent Systems, 31(1), 35–50.
• Alkan, R., Yıldız, A., Ve Gök, M. (2025), Xgboost Tabanlı Pozisyon Tahmini Modeli İle Borsa İstanbul’da İşlem Stratejisi Geliştirme, Finansal Teknolojiler Ve Analitik Dergisi, 12(1), 45–60.
• Aras, S. (2020), Using Technical İndicators To Predict Stock Price İndex Movements By Machine Learning Techniques. In E. Sarikaya (Ed.), Theory And Research İn Social, Human And Administrative Sciences Iı (S. 249–274). Gece Publishing.
• Bao, W., Yue, J., Ve Rao, Y. (2017), A Deep Learning Framework For Financial Time Series Using Stacked Autoencoders And Long-Short Term Memory, Plos One, 12(7), E0180944.
• Basak, S., Kar, S., Saha, S., Khaıdem, L., Ve Dey, S. R. (2019), Predicting The Direction Of Stock Market Prices Using Tree-Based Classifiers, The North American Journal Of Economics And Finance, 47, 552–567.
• Breıman, L. (2001), Random Forests, Machine Learning, 45(1), 5–32.
• Büyükkör, Y., Ve Doğan, S. (2024), Borsa Endeks Yönünün Ağaç Tabanlı Topluluk Makine Öğrenmesi Yöntemleri İle Tahmini: Bist-100 Örneği, Bingöl Üniversitesi Sosyal Bilimler Enstitüsü Dergisi, 27, 324–335.
• Chen, A. S., Leung, M. T., Ve Daouk, H. (2018), Application Of Neural Networks To An Emerging Financial Market: Forecasting And Trading The Taiwan Stock Index, Computers & Operations Research, 30(6), 901–923.
• Chen, T. Ve Guestrın, C. (2016), Xgboost: A Scalable Tree Boosting System, Proceedings Of The 22nd Acm Sıgkdd International Conference On Knowledge Discovery And Data Mining, 785–794.
• Fama, E. F. (1995), Random Walks İn Stock Market Prices, Financial Analysts Journal, 51(1), 75–80.
• Fang, Z., Ma, X., Pan, H., Yang, G., Ve Arce, G. R. (2023), Movement Forecasting Of Financial Time Series Based On Adaptive Lstm-Bn Network, Expert Systems With Applications, 213, 119207.
• Fıscher, T. Ve Krauss, C. (2018), Deep Learning With Long Short-Term Memory Networks For Financial Market Predictions, European Journal Of Operational Research, 270(2), 654–669.
• Gumelar, A. B., Setyorını, H., Adı, D. P., Nılowardono, S., Latıpah, Ve Wıdodo, A. (2022), Boosting The Accuracy Of Stock Market Prediction Using Xgboost And Long Short-Term Memory, 2022 International Conference On Information Technology Research And Innovation (Icıtrı), 1–6.
• Gupta, A., Sıngh, R., Ve Awasthı, A. (2025), Comparative Analysis Of Lstm And Gru Networks For Sentiment-Enhanced Stock Price Prediction İn Indian Markets, Journal Of Computational Finance And Aı, 4(1), 22–38.
• Hu, H., Tang, L., Zhang, S., Ve Wang, H. (2018), Predicting The Direction Of Stock Markets Using Optimized Neural Networks With Google Trends, Neurocomputing, 285, 188–195.
• Jang, H., Ve Lee, D. (2024), Predicting Stock Market Returns Using Ensemble Learning Methods: Evidence From The Kospı, Asian Journal Of Financial Research, 19(3), 101–115.
• Kara, Y., Boyacıoğlu, M. A., Ve Baykan, Ö. K. (2011), Predicting Direction Of Stock Price İndex Movement Using Artificial Neural Networks And Support Vector Machines: The Sample Of The Istanbul Stock Exchange, Expert Systems With Applications, 38(5), 5311–5319.
• Kılıç, A., Ve Yılmaz, H. (2023), Rf Ve Xgboost Algoritmalarıyla Borsa İstanbul’da Yön Tahmini, Finansal Uygulamalar Ve Strateji Dergisi, 15(2), 85–97.
• Krollner, B., Vanstone, B., Ve Fınnıe, G. (2010), Financial Time Series Forecasting With Machine Learning Techniques: A Survey, Esann 2010 Proceedings, 1–6.
• Lebaron, B. (1992), Forecast İmprovements Using A Volatility-Based Model Of Stock Returns, International Journal Of Forecasting, 8(2), 255–271.
• Mıshra, O., Joshı, U., Patıl, H., Ve Dongardıve, J. (2024), Comparing Lstm And Random Forests For Stock Price Movement Forecasting, International Journal For Research Trends And Innovation, 9(1), 271–279.
• Omar, A. B., Huang, S., Salameh, A. A., Khurram, H., Ve Fareed, M. (2022), Stock Market Forecasting Using The Random Forest And Deep Neural Network Models Before And During The Covıd-19 Period, Frontiers İn Environmental Science, 10, 917047.
• Omar, M. N., Alı, M., Ve Rahman, A. (2022), Predicting Stock İndex Movements Using Random Forest And Arıma: A Comparative Study, Journal Of Forecasting, 41(5), 897–913.
• Ouf, A., Easa, N., Farag, M., Ve Solıman, F. (2024), A Deep Learning Based Lstm For Stock Price Prediction Using Sentiment Analysis, International Journal Of Advanced Computer Science And Applications (Ijacsa), 15(12), 188–198
• Patel, J., Shah, S., Thakkar, P., Ve Kotecha, K. (2015), Predicting Stock And Stock Price İndex Movement Using Trend Deterministic Data Preparation And Machine Learning Techniques, Expert Systems With Applications, 42(1), 259–268.
• Qıan, Z. (2024), Comparison Of Machine Learning Models For Stock Price Direction Prediction: A Case Study Of S&P 500 And Csı 300, Proceedings Of The 13th International Conference On Data Science, Technology And Applications (Data 2024), 252–261.
• Rathı, S., Patıl, K., Ve Ingle, V. (2024), Portfolio Volatility Prediction Using Machine Learning Algorithms, Proceedings Of The 16th International Conference On Agents And Artificial Intelligence (Icaart 2024), 398–405.
• Smıth, J., Brown, L., Ve Taylor, R. (2025), Aı İntegration İn Financial Services: A Systematic Review Of Trends And Regulatory Challenges, Humanities And Social Sciences Communications, 12(1), 4850
• Tay, F. E. H., Ve Cao, L. (2001), Application Of Support Vector Machines İn Financial Time Series Forecasting, Omega, 29(4), 309–317.
• Uhunmwangho, O. R. (2024), A Comparative Study Of Machine Learning Algorithms For Stock Market Prediction: The Case Of Microsoft Stock, Mujast - Midwest Journal Of Applied Science & Technology, 4(2), 52–65.
• Vuong, P. H., Dat, T. T., Maı, T. K., Uyen, P. H., Ve Bao, P. T. (2022), Stock-Price Forecasting Based On Xgboost And Lstm, Computer Systems Science And Engineering, 40(1), 237–246.
• Vuong, H. N., Do, H. H., Ve Pham, M. T. (2022), Hybrid Feature Extraction And Machine Learning Models For Financial Time Series Forecasting, Expert Systems With Applications, 195, 116595.
• Zhang, X., Aggarwal, C. C., Ve Qı, G.-J. (2020), Stock Price Prediction Via Discovering Multi-Frequency Trading Patterns, Proceedings Of The 26th Acm Sıgkdd International Conference On Knowledge Discovery & Data Mining, 2723–2733.
• Zhang, X., Ve Yu, B. (2023), Attention-Enhanced Lstm For Stock Volatility Prediction Under Turbulent Market Conditions, Journal Of Financial Data Science, 5(2), 67–82.
• Zheng, J., Xın, D., Cheng, Q., Tıan, M., Ve Yang, L. (2024), The Random Forest Model For Analyzing And Forecasting The Us Stock Market İn The Context Of Smart Finance, Proceedings Of The 3rd International Academic Conference On Blockchain, Information Technology And Smart Finance (Icbıs 2024), 82–90.
• Zheng, Q., Lı, Z., Ve Huang, Y. (2024), Evaluating Technical İndicators İn Stock Market Prediction Using Ensemble Learning, Computational Economics And Finance Review, 32(1), 120–139.