DEEP LEARNING IN FINANCIAL ANALYSIS: APPLICATION OF LSTM AND GANs FOR STOCK PRICE PREDICTION
DOI:
https://doi.org/10.34185/1991-7848.itmm.2025.01.106Keywords:
Time series forecasting, SMA, EMA, ARIMA, LSTM, GANsAbstract
Stock price prediction is a crucial aspect of financial analytics, helping investors make informed decisions. This study examines traditional forecasting methods, such as technical analysis (moving averages SMA, EMA) and statistical models (ARIMA, exponential smoothing). Their advantages and limitations are analyzed, particularly the challenges in capturing complex market patterns. To improve prediction accuracy, the use of modern machine learning approaches is proposed, specifically Long Short-Term Memory (LSTM) networks and Generative Adversarial Networks (GANs). The GAN architecture and its ability to model market dynamics even with limited historical data are described. The research is based on real stock market data (AAPL stock prices), and the results are compared with ARIMA and LSTM methods, confirming the effectiveness of the proposed approach.
References
Pertsev Y. O., Korotka L.I. Comparative analysis of traditional statistical methods and the LSTM neural network model. System technologies. 2025. №1(156). P. 65-77.
Pertsev Y. O., Korotka L. I. Porivniannia neironnykh merezh RNN ta LSTM typu pry prohnozuvanni tsin na fondovomu rynku. Materialy VIII Mizhnarodnoi naukovo-tekhnichnoi konferentsii kompiuterne modeliuvannia ta optymizatsiia skladnykh system (1-3 lystopada 2023 roku m. Dnipro, Ukraina). S. 124-127.
Korotka L., Klevzhyts D., Shvydko D. Use of generative-adversarial networks when creating content. Artificial intelligence. National Academy of Sciences of Ukraine In-stitute of Artificial Intelligence Problems MES of Ukraine and NAS of Ukraine. 2024 № 2 (99). P. 32-47. https://doi.org/10.15407/jai2024.02.032
K. Zhang, G. Zhong, J. Dong, S. Wang, and Y. Wang, “Stock Market Prediction Based on Generative Adversarial Network,” Procedia Computer Science, vol. 147, pp. 400–406, 2019.
