NEURAL NETWORK PREDICTION OF PRICES ON THE STOCK MARKET
DOI:
https://doi.org/10.34185/1991-7848.itmm.2023.01.085Keywords:
neural networks, forecasting, stock market.Abstract
In the modern world of financial markets are showing more and more interest in systems that can accurately and efficiently predict the future price of financial instruments. Information technologies that exist at the moment make it possible to analyze and use highly loaded systems such as neural networks to predict the financial performance of a company. This article discusses the use of neural networks to predict the price of shares in the financial market. The possibilities of neural networks for predicting stock prices are considered due to the fact that neural networks have many hidden blocks that allow the model to adapt to complex relationships between company indicators and its stock price. An example of a RNN neural network is given that can work with sequential data such as hourly series. For an example of building a model, Apple was chosen as one of the largest companies that is included in the S&P 500 list of the most influential companies in the American market.
References
Kholoshnia D.M., Korotka L.I. Informatsiina pidsystema neiromerezhevoho prohnozuvannia finansovykh sytuatsii na valiutnomu rynku / VIII Mizhnarodna naukovo-tekhnichna konferentsiia studentiv, aspirantiv ta molodykh vchenykh «Khimiia ta suchasni tekhnolohii». Tezy dopovidei. V Tom (27-29 kvitnia 2017 Dnipro). – 2017. S. 26. 2. Ian Goodfellow, Yoshua Bengio, Aaron Courville Deep Learning [Elektronnyy̆ resurs] URL: https://www.deeplearningbook.org/ 3. Korotka L.I. Funktsionalna pidsystema ratsionalnoho vyboru arkhitektury neironnoi merezhi /L.I. Korotka // Visnyk Khersonskoho natsionalnoho tekhnichnoho universytetu 3(62), Tom I. (Fundamentalni nauky). – 2017. S. 55-59.
