AI-OPTIMIZED FUZZY CONVOLUTIONAL PREDICTION FOR FINANCIAL TIME SERIES

Furkan Keskin; Ozge Cagcag Yolcu; Sümeyye Polater

doi:10.37880/cumuiibf.1885701

EN TR

AI-OPTIMIZED FUZZY CONVOLUTIONAL PREDICTION FOR FINANCIAL TIME SERIES

Abstract

A deep convolutional prediction approach for financial time series is introduced in which fuzzy-cluster memberships and lagged observations are provided as inputs to a GA-tuned 1D convolutional network. Fuzzy C-Means clustering is fitted on the training segment to obtain memberships, whose nonlinear transforms are concatenated with scaled autoregressive lags to form a single feature block processed by the network. Hyperparameters—including lag order, number of clusters, fuzzifier, convolutional kernel size, filter counts, and mini-batch size—are selected through a genetic algorithm that minimizes validation RMSE. The proposed AI-Optimized Fuzzy Convolutional Prediction approach, incorporating fuzzy memberships and autoregressive lags, is evaluated on TAIEX datasets comprising 16 series. Across these datasets, lower error rates than state-of-the-art baselines are obtained, and close agreement between predictions and realized values is indicated by regression diagnostics and visual graphics. An intercept-free regression of observations on forecasts yielded slopes and determination coefficients consistently near one, and scatter/residual plots exhibited limited dispersion and no salient error structure. The results suggest that integrating fuzzy membership–based feature construction with GA-tuned convolutional modelling yields accurate predictions on financial time series.

Keywords

FİNANSAL ZAMAN SERİLERİ İÇİN YAPAY ZEKA İLE OPTİMİZE EDİLMİŞ BULANIK EVRİŞİMSEL TAHMİN

Öz

Finansal zaman serileri için, bulanık küme üyelikleri ve gecikmeli gözlemlerin genetik algoritma ile ayarlanan tek boyutlu (1D) bir evrişimsel ağa girdi olarak verildiği derin evrişimsel bir öngörü yaklaşımı sunulmaktadır. Eğitim bölümüne Bulanık C-Ortalamalar uygulanarak üyelikler elde edilmekte; bu üyeliklerin doğrusal olmayan dönüşümleri, ölçeklenmiş otoregresif gecikmelerle birleştirilerek ağ tarafından işlenen tek bir özellik bloğu oluşturulmaktadır. Gecikme derecesi, küme sayısı, bulanıklık indeksi, evrişim çekirdeği boyutu, filtre sayıları ve mini-yığın boyutu dâhil hiperparametreler, doğrulama setleri üzerinden hata kareler ortalamasını en aza indiren bir genetik algoritma ile seçilmektedir. Bulanık üyelikler ve otoregresif gecikmelerle oluşturulan Yapay Zekâ ile Optimize Edilen Bulanık Evrişimsel Tahmin yaklaşımı, 16 farklı Tayvan Borsası zaman serisi ne uygulanmıştır. Bu veri setleri için, güncel yöntemlere kıyasla daha düşük hata oranları elde edilmiştir. Gözlemler üzerine, sabit terim içermeyen bir doğrusal regresyon kurulduğunda, eğim ve belirleme katsayılarının tutarlı biçimde bire yakın olduğu; saçılım ve artık grafiklerinde sınırlı yayılım ve belirgin bir hata yapısının bulunmadığı gözlenmiştir. Bulgular, önerilen tahmin modelinin oldukça başarılı tahminler sağladığını ortaya koymaktadır.

Anahtar Kelimeler

Destekleyen Kurum

Marmara Üniversitesi

Proje Numarası

FYL-2024-11360

Teşekkür

Bu çalışma, Marmara Üniversitesi Bilimsel Araştırma Projeleri Komitesi (BAPKO) tarafından FYL-2024-11360 numaralı Yüksek Lisans Tezi Projesi kapsamında desteklenmiştir.

Kaynakça

Aktoprak M.R. & Cagcag Yolcu O. (2025). A New Approach for Time Series Prediction: Fuzzy Regression Network Functions. International Journal of Advances in Engineering and Pure Sciences, 37(1), 36-52. doi:10.7240/jeps.1573839.
Aladag C.H., Basaran M.A., Egrioglu E., Yolcu U. & Uslu V.R. (2009). Forecasting in High Order Fuzzy Times Series by Using Neural Networks to Define Fuzzy Relations. Expert Systems with Applications, 36(3 PART 1). doi:10.1016/j.eswa.2008.04.001.
Aladag C.H., Yolcu U. & Egrioglu E. (2010a). A High Order Fuzzy Time Series Forecasting Model Based on Adaptive Expectation and Artificial Neural Networks. Mathematics and Computers in Simulation, 81(4). doi:10.1016/j.matcom.2010.09.011.
Aladag C.H., Yolcu U. & Egrioglu E. (2010b). A High Order Fuzzy Time Series Forecasting Model Based on Adaptive Expectation and Artificial Neural Networks. Mathematics and Computers in Simulation, 81(4). doi:10.1016/j.matcom.2010.09.011.
Aladag C.H., Yolcu U., Egrioglu E. & Turksen I.B. (2016). Type-1 Fuzzy Time Series Function Method Based on Binary Particle Swarm Optimisation. International Journal of Data Analysis Techniques and Strategies, 8(1), 2-13. https://dx.doi.org/10.154/IJDATS.2016.075970
Alpaslan F., Cagcag O., Aladag C.H., Yolcu U. & Egrioglu E. (2012). A Novel Seasonal Fuzzy Time Series Method. Hacettepe Journal of Mathematics and Statistics, 41(3).
Bas E., Egrioglu E., Aladag C.H. & Yolcu U. (2015). Fuzzy-Time-Series Network Used to Forecast Linear and Nonlinear Time Series. Applied Intelligence, 43(2). doi:10.1007/s10489-015-0647-0.
Bas E., Egrioglu E., Yolcu U. & Grosan C. (2019a). Type 1 Fuzzy Function Approach Based on Ridge Regression for Forecasting. Granular Computing, 4(4). doi:10.1007/s41066-018-0115-4.

Bas E., Egrioglu E., Yolcu U. & Grosan C. (2019b). Type 1 Fuzzy Function Approach Based on Ridge Regression for Forecasting. Granular Computing, 4(4). doi:10.1007/s41066-018-0115-4.
Box G.E.P., Jenkins G.M. & Reinsel G.C. (2013). Time Series Analysis: Forecasting and Control: Fourth Edition.
Cagcag Yolcu O. & Alpaslan F. (2018). Prediction of TAIEX Based on Hybrid Fuzzy Time Series Model with Single Optimization Process. Applied Soft Computing Journal, 66. doi:10.1016/j.asoc.2018.02.007.
Cagcag Yolcu O., Bas E., Egrioglu E. & Yolcu U. (2020). A New Intuitionistic Fuzzy Functions Approach Based on Hesitation Margin for Time-Series Prediction. Soft Computing, 24(11). doi:10.1007/s00500-019-04432-2.
Cai Q., Zhang D., Zheng W. & Leung S.C.H. (2015). A New Fuzzy Time Series Forecasting Model Combined with Ant Colony Optimization and Auto-Regression. Knowledge-Based Systems, 74. doi:10.1016/j.knosys.2014.11.003.
Chen S.M. (1996a). Forecasting Enrollments Based on Fuzzy Time Series. Fuzzy Sets and Systems, 81(3). doi:10.1016/0165-0114(95)00220-0.
Chen S.M. (1996b). Forecasting Enrollments Based on Fuzzy Time Series. Fuzzy Sets and Systems, 81(3). doi:10.1016/0165-0114(95)00220-0.
Chen S.M. (2002). Forecasting Enrollments Based on High-Order Fuzzy Time Series. Cybernetics and Systems, 33(1). doi:10.1080/019697202753306479.
Chen S.M. & Chang Y.C. (2010). Multi-Variable Fuzzy Forecasting Based on Fuzzy Clustering and Fuzzy Rule Interpolation Techniques. Information Sciences, 180(24). doi:10.1016/j.ins.2010.08.026.
Chen S.M. & Chen C.D. (2011). TAIEX Forecasting Based on Fuzzy Time Series and Fuzzy Variation Groups. IEEE Transactions on Fuzzy Systems, 19(1). doi:10.1109/TFUZZ.2010.2073712.
Chen S.M. & Chen S.W. (2015). Fuzzy Forecasting Based on Two-Factors Second-Order Fuzzy-Trend Logical Relationship Groups and the Probabilities of Trends of Fuzzy Logical Relationships. IEEE Transactions on Cybernetics, 45(3). doi:10.1109/TCYB.2014.2326888.
Chen S.M., Chu H.P. & Sheu T.W. (2012). TAIEX Forecasting Using Fuzzy Time Series and Automatically Generated Weights of Multiple Factors. IEEE Transactions on Systems, Man, and Cybernetics Part A:Systems and Humans, 42(6). doi:10.1109/TSMCA.2012.2190399.
Chen Y.S., Cheng C.H., Chiu C.L. & Huang S.T. (2016). A Study of ANFIS-Based Multi-Factor Time Series Models for Forecasting Stock Index. Applied Intelligence, 45(2). doi:10.1007/s10489-016-0760-8.
Cheng C., Chen T., Teoh H. & Chiang C. (2008). Fuzzy Time-Series Based on Adaptive Expectation Model for TAIEX Forecasting. Expert Systems with Applications, 34(2), 1126-1132. doi:10.1016/j.eswa.2006.12.021.
Cheng S.H., Chen S.M. & Jian W.S. (2016). Fuzzy Time Series Forecasting Based on Fuzzy Logical Relationships and Similarity Measures. Information Sciences, 327. doi:10.1016/j.ins.2015.08.024.
Dalar A.Z., Egrioglu E., Yolcu U. & Aladag C.H. (2017). Fuzzy Functions Approach for Time Series Forecasting. In A.Z. Dalar, E. Egrioglu, U. Yolcu & C.H. Aladag (Eds.), Advances in Time Series Forecasting Volume: 2 (pp. 144-155).
Dong Q. & Ma X. (2021). Enhanced Fuzzy Time Series Forecasting Model Based on Hesitant Differential Fuzzy Sets and Error Learning. Expert Systems with Applications, 166. doi:10.1016/j.eswa.2020.114056.
Egrioglu E. (2014). PSO-Based High Order Time Invariant Fuzzy Time Series Method: Application to Stock Exchange Data. Economic Modelling, 38. doi:10.1016/j.econmod.2014.02.017.
Egrioglu E., Aladag C.H., Yolcu U. & Bas E. (2015). A New Adaptive Network Based Fuzzy Inference System for Time Series Forecasting. Aloy Journal of Soft Computing and Applications, 2(1).
Egrioglu E., Aladag C.H., Yolcu U., Uslu V.R. & Basaran M.A. (2009). A New Approach Based on Artificial Neural Networks for High Order Multivariate Fuzzy Time Series. Expert Systems with Applications, 36(7). doi:10.1016/j.eswa.2009.02.057.
Egrioglu E., Aladag C.H., Yolcu U., Uslu V.R. & Basaran M.A. (2010). Finding an Optimal Interval Length in High Order Fuzzy Time Series. Expert Systems with Applications, 37(7). doi:10.1016/j.eswa.2009.12.006.
Fatima S.S.W. & Rahimi A. (2024). A Review of Time-Series Forecasting Algorithms for Industrial Manufacturing Systems. Machines, 12(6), 380. doi:10.3390/machines12060380.
Hsu L.Y., Horng S.J., Kao T.W., Chen Y.H., Run R.S., Chen R.J., Lai J.L. & Kuo I.H. (2010). Temperature Prediction and TAIFEX Forecasting Based on Fuzzy Relationships and MTPSO Techniques. Expert Systems with Applications, 37(4). doi:10.1016/j.eswa.2009.09.015.
Huarng K.H., Yu T.H.K. & Hsu Y.W. (2007). A Multivariate Heuristic Model for Fuzzy Time-Series Forecasting. IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics, 37(4). doi:10.1109/TSMCB.2006.890303.
Huarng K. (2001). Effective Lengths of Intervals to Improve Forecasting in Fuzzy Time Series. Fuzzy Sets and Systems, 123(3). doi:10.1016/S0165-0114(00)00057-9.
Huarng K. & Yu T.H.K. (2006). The Application of Neural Networks to Forecast Fuzzy Time Series. Physica A: Statistical Mechanics and Its Applications, 363(2). doi:10.1016/j.physa.2005.08.014.
Hyndman R.J. & Athanasopoulos G. (2018). 3 . 4 Evaluating Forecast Accuracy Training and Test Sets. Forecasting: Principles and Practice (Online-Version, Kontinuierlich Aktualisiert).
Jang J.S.R. (1993a). ANFIS: Adaptive-Network-Based Fuzzy Inference System. IEEE Transactions on Systems, Man and Cybernetics, 23(3). doi:10.1109/21.256541.
Jang J.S.R. (1993b). ANFIS: Adaptive-Network-Based Fuzzy Inference System. IEEE Transactions on Systems, Man and Cybernetics, 23(3). doi:10.1109/21.256541.
Kirisci M. & Cagcag Yolcu O. (2022). A New CNN-Based Model for Financial Time Series: TAIEX and FTSE Stocks Forecasting. Neural Processing Letters, 54(4). doi:10.1007/s11063-022-10767-z.
Kocak C., Dalar A.Z., Cagcag Yolcu O., Bas E. & Egrioglu E. (2020). A New Fuzzy Time Series Method Based on an ARMA-Type Recurrent Pi-Sigma Artificial Neural Network. Soft Computing, 24(11). doi:10.1007/s00500-019-04506-1.
Kontopoulou V.I., Panagopoulos A.D., Kakkos I. & Matsopoulos G.K. (2023). A Review of ARIMA vs. Machine Learning Approaches for Time Series Forecasting in Data Driven Networks. Future Internet, 15(8).
Nor Syazwina Binti Mohd Hanafiah, Nor Hayati Binti Shafii, Nur Fatihah Binti Fauzi, Diana Sirmayunie Mohd Nasir & Nor Azriani Mohamad Nor (2022). Comparison of Fuzzy Time Series and ARIMA Model for Predicting Stock Prices. Journal of Computing Research and Innovation, 7(2). doi:10.24191/jcrinn.v7i2.332.
Polater S., Yolcu U., Keskin F. & Cagcag Yolcu O. (2026). Picture Fuzzy C-Means–Based Ensemble of Forecasting Functions for Financial Time Series Forecasting. International Journal of Advances in Engineering and Pure Sciences, 38(1), 199-209. doi:10.7240/jeps.1818060.
Sadaei H.J., Enayatifar R., Lee M.H. & Mahmud M. (2016). A Hybrid Model Based on Differential Fuzzy Logic Relationships and Imperialist Competitive Algorithm for Stock Market Forecasting. Applied Soft Computing Journal, 40. doi:10.1016/j.asoc.2015.11.026.
Sarıca B., Eğrioğlu E. & Aşıkgil B. (2018). A New Hybrid Method for Time Series Forecasting: AR–ANFIS. Neural Computing and Applications, 29(3). doi:10.1007/s00521-016-2475-5.
Song Q. & Chissom B.S. (1993a). Forecasting Enrollments with Fuzzy Time Series - Part I. Fuzzy Sets and Systems, 54(1). doi:10.1016/0165-0114(93)90355-L.
Song Q. & Chissom B.S. (1993b). Fuzzy Time Series and Its Models. Fuzzy Sets and Systems, 54(3). doi:10.1016/0165-0114(93)90372-O.
Song Q. & Chissom B.S. (1993c). Fuzzy Time Series and Its Models. Fuzzy Sets and Systems, 54(3). doi:10.1016/0165-0114(93)90372-O.
Song Q. & Chissom B.S. (1994). Forecasting Enrollments with Fuzzy Time Series - Part II. Fuzzy Sets and Systems, 62(1). doi:10.1016/0165-0114(94)90067-1.
Sun G. & Deng S. (2025). Financial Time Series Forecasting: A Comparison Between Traditional Methods and AI-Driven Techniques. Journal of Computer, Signal, and System Research, 2(2), 86-93. doi:10.71222/339b9812.
Tak N. (2022). A Novel ARMA Type Possibilistic Fuzzy Forecasting Functions Based on Grey-Wolf Optimizer (ARMA-PFFs). Computational Economics, 59(4). doi:10.1007/s10614-021-10132-7.
Tak N., Evren A.A., Tez M. & Egrioglu E. (2018a). Recurrent Type-1 Fuzzy Functions Approach for Time Series Forecasting. Applied Intelligence, 48(1). doi:10.1007/s10489-017-0962-8.
Tak N., Evren A.A., Tez M. & Egrioglu E. (2018b). Recurrent Type-1 Fuzzy Functions Approach for Time Series Forecasting. Applied Intelligence, 48(1). doi:10.1007/s10489-017-0962-8.
Tak N. & İnan D. (2022). Type-1 Fuzzy Forecasting Functions with Elastic Net Regularization. Expert Systems with Applications, 199. doi:10.1016/j.eswa.2022.116916.
Türkşen I.B. (2008). Fuzzy Functions with LSE. Applied Soft Computing Journal, 8(3). doi:10.1016/j.asoc.2007.12.004.
Wang C.C. (2011). A Comparison Study between Fuzzy Time Series Model and ARIMA Model for Forecasting Taiwan Export. Expert Systems with Applications, 38(8). doi:10.1016/j.eswa.2011.01.015.
Yolcu U., Aladag C.H., Egrioglu E. & Uslu V.R. (2013). Time-Series Forecasting with a Novel Fuzzy Time-Series Approach: An Example for Istanbul Stock Market. Journal of Statistical Computation and Simulation, 83(4). doi:10.1080/00949655.2011.630000.
Yu T.H.K. & Huarng K.H. (2008a). A Bivariate Fuzzy Time Series Model to Forecast the TAIEX. Expert Systems with Applications, 34(4). doi:10.1016/j.eswa.2007.05.016.
Yu T.H.K. & Huarng K.H. (2008b). A Bivariate Fuzzy Time Series Model to Forecast the TAIEX. Expert Systems with Applications, 34(4). doi:10.1016/j.eswa.2007.05.016.
Zadeh L.A. (1965). Fuzzy Sets. Information and Control, 8(3). doi:10.1016/S0019-9958(65)90241-X.

Ayrıntılar

Birincil Dil

İngilizce

Konular

Zaman Serileri Analizi , İstatistik (Diğer)

Bölüm

Araştırma Makalesi

Yazarlar

Furkan Keskin
0000-0002-8178-4227
Türkiye

Ozge Cagcag Yolcu ^*
0000-0003-3339-9313
Türkiye

Sümeyye Polater
0009-0003-6010-7387
Türkiye

Yayımlanma Tarihi

26 Nisan 2026

Gönderilme Tarihi

10 Şubat 2026

Kabul Tarihi

24 Nisan 2026

Yayımlandığı Sayı

Yıl 2026 Cilt: 27 Sayı: 2

DOI

https://doi.org/10.37880/cumuiibf.1885701

IZ

https://izlik.org/JA67UT23AW

Kaynak Göster

RIS / Bibtex

APA

Keskin, F., Cagcag Yolcu, O., & Polater, S. (2026). AI-OPTIMIZED FUZZY CONVOLUTIONAL PREDICTION FOR FINANCIAL TIME SERIES. Cumhuriyet Üniversitesi İktisadi ve İdari Bilimler Dergisi, 27(2), 816-838. https://doi.org/10.37880/cumuiibf.1885701