ANALYSIS AND ADAPTATION OF GENERATIVE ADVERSARIAL NETWORKS FOR FINANCIAL MARKETS
Keywords:
generative adversarial networks, stock market forecasting, deep learning, LSTM, CNN, synthetic data, financial time series, machine learning, RMSE, technical indicators, neural networks, Apple IncAbstract
The article investigates the applicability of Generative Adversarial Networks (GAN) for financial forecasting, focusing on the prediction of stock closing prices. A novel GAN architecture is proposed, where the generator is implemented using a Long Short-Term Memory (LSTM) network, and the discriminator is designed as a Convolutional Neural Network (CNN).The primary goal is to synthesize realistic financial time series based on market features, which can be used to train and improve predictive models. The model is trained and tested on a five-year historical dataset (2020–2025) for Apple Inc., incorporating more than 30 input features, including stock price parameters (Open, High, Low, Close, Volume), technical indicators (MACD, EMA, Bollinger Bands), global stock indices, sentiment-related components, and harmonic transformations. A comparative analysis of the GAN-based model and a classical LSTM model was conducted using Root Mean Square Error (RMSE) as the key performance metric. The results demonstrate that the proposed GAN model achieves superior accuracy and robustness, particularly in identifying nonlinear patterns and complex dependencies within financial time series.Moreover, the use of synthetic data generated by the GAN significantly enhances the training process, leading to better generalization in volatile markets.The study shows that the GAN architecture is capable of capturing both upward and downward market trends and adjusting to short- and long-term temporal patterns. This adaptability, coupled with the modular design of the network, makes the proposed solution promising for developing AI-powered tools in financial analytics. The research outcomes can be leveraged by traders, investors, financial analysts, and fintech developers aiming to incorporate deep learning-based forecasting methods into practical investment decision-making processes.
References
Alberto, R., & Romero, C. (2019). Generative Adversarial Network for Stock Market price Prediction. URL: https://cs230.stanford.edu/projects_fall_2019/reports/26259829.pdf
Lin, H., Chen, C., Huang, G. & Jafari, A. (2021). Stock price prediction using Generative Adversarial Networks. Journal of Computer Science, 17(3), 188–196. DOI: https://doi.org/10.3844/jcssp.2021.188.196
Staffini A (2022) Stock Price Forecasting by a Deep Convolutional Generative Adversarial Network. Front. Artif. Intell. 5:837596. DOI: https://doi.org/10.3389/ frai.2022.837596
Santoro, D., Grilli, L. (2022) Generative adversarial network to evaluate quantity of information in financial markets. Neural Comput. Appl. 34, 17473–17490. DOI: https://doi.org/10.1007/s00521-022-07401-3
Hutto, C., & Gilbert, E. (2014). VADER: A Parsimonious Rule-Based Model for Sentiment Analysis of Social Media Text. Proceedings of the International AAAI Conference on Web and Social Media, 8(1), 216–225. DOI: https://doi.org/10.1609/ icwsm.v8i1.14550
M. J. R. Foo and C. S. Pun, (2022) “Stock Movement Prediction with Social Sentiments and Interactional Data: Integrating NLP and Bayesian Frameworks”, 2022 4th International Conference on Natural Language Processing (ICNLP), Xi’an, China, 2022, pp. 530–536, DOI: https://doi.org/10.1109/ICNLP55136.2022.00097.
Cristescu, M. P., Nerisanu, R. A., Mara, D. A., & Oprea, S.-V. (2022). Using Market News Sentiment Analysis for Stock Market Prediction. Mathematics, 10(22), 4255. DOI: https://doi.org/10.3390/math10224255
Zhang Y, Li J, Wang H and Choi S-CT (2021) Sentiment-Guided Adversarial Learning for Stock Price Prediction. Front. Appl. Math. Stat. 7:601105. DOI: https://doi.org/ 10.3389/fams.2021.601105
S. Mukherjee, B. Sadhukhan, N. Sarkar, D. Roy, and S. De, (2023) “Stock market prediction using deep learning algorithms,” CAAI Transactions on Intelligence Technology, vol. 8, no. 1, pp. 82–94 DOI: https://doi.org/10.1049/cit2.12059
E. Naresh, B. J. Ananda, K. S. Keerthi and M. R. Tejonidhi, (2022) “Predicting the Stock Price Using Natural Language Processing and Random Forest Regressor”, 2022 IEEE International Conference on Data Science and Information System (ICDSIS), Hassan, India, 2022, pp. 1-5, DOI: https://doi.org/10.1109/ICDSIS55133.2022.9915940
Перцев Ю. О., Коротка Л. І. (2025) Порівняльний аналіз традиційних статистичних методів та нейромережевої моделі LSTM. Системні технології. № 1(156). С. 65–77.
Перцев Ю. О., Коротка Л. І. (2024) Інноваційний підхід у прогнозуванні часових рядів: від традиційних методів до новаторської моделі TIMESFM. International scientific and technical conference Information Тechnologies in Metallurgy and Machine building (Україна, м. Дніпро, 10-11 квітня 2024 року), Дніпро : ITMM 2024, С. 434–439.
Korotka L., Klevzhyts D., Shvydko D. (2024) Use of generative-adversarial networks when creating content. Artificial intelligence. National Academy of Sciences of Ukraine Institute of Artificial Intelligence Problems MES of Ukraine and NAS of Ukraine. 2024 № 2 (99). P. 32–47.
Перцев Ю. О., Коротка Л. І. (2023) Нейромережеве прогнозування цін на фондовому ринку. International scientific and technical conference Information Тechnologies in Metallurgy and Machine building (Україна, м. Дніпро, 22 березня 2023 року), Дніпро : ITMM 2023, С. 314–317.
Перцев Ю. О., Коротка Л. І. (2023) Порівняння нейронних мереж RNN та LSTM типу при прогнозуванні цін на фондовому ринку. Матеріали VІІІ Міжнародної науково-технічної конференції комп’ютерне моделювання та оптимізація складних систем (Україна, м. Дніпро, 1–3 листопада 2023 року). Дніпро : КМОСС-2023 С. 124–127.
Nadeem Malibari, Iyad Katib and Rashid Mehmood, (2021) “Predicting Stock Closing Prices in Emerging Markets with Transformer Neural Networks: The Saudi Stock Exchange Case” International Journal of Advanced Computer Science and Applications(IJACSA), vol. 12, no. 12. DOI: http://dx.doi.org/10.14569/IJACSA.2021.01212106
Guennec, A. L., Malinowski, S. & Tavenard, R. (2016). Data Augmentation for Time Series Classification using Convolutional Neural Networks URL: https://halshs.archives-ouvertes.fr/halshs-01357973/
Cui, Zhicheng & Chen, Wenlin. (2016). Multi-Scale Convolutional Neural Networks for Time Series Classification. DOI: https://doi.org/10.48550/arXiv.1603.06995.
Kim T, Kim HY (2019) Forecasting stock prices with a feature fusion LSTM- CNN model using different representations of the same data. PLoS ONE 14(2): e0212320. DOI: https://doi.org/10.1371/journal.pone.0212320
Lu, Wenjie & Li, Jiazheng & Li, Yifan & Sun, Aijun & Wang, Jingyang. (2020). A CNN-LSTM-based model to forecast stock prices. Complexity. 2020. 1–10. DOI: https://doi.org/10.1155/2020/6622927
Kirisci, Melih & Cagcag Yolcu, Ozge. (2022). A New CNN-Based Model for Financial Time Series: TAIEX and FTSE Stocks Forecasting. Neural Processing Letters. 54. 10.1007/s11063-022-10767-z.
Mehtab, S., Sen, J. (2022). Analysis and Forecasting of Financial Time Series Using CNN and LSTM-Based Deep Learning Models. In: Sahoo, J. P., Tripathy, A. K., Mohanty, M., Li, KC., Nayak, A. K. (eds) Advances in Distributed Computing and Machine Learning. Lecture Notes in Networks and Systems, vol. 302. Springer, Singapore. DOI: https://doi.org/10.1007/978-981-16-4807-6_39
