STUDY OF NEURAL NETWORKS FOR FORECASTING THE VALUE OF COMPANY SHARES IN AN UNSTABLE ECONOMY

These studies deal with analysis and selection of neural networks various architectures hybrid models, which include networks, to predict the market value shares in stock a country that is process unstable development. Analysis forecasting such markets cannot be carried out using classical methods. The relevance research topic due need develop software systems implement algorithmic support for predicting Ukraine. introduction circuit investment decisionmaking companies are interested increasing information transparency Ukrainian will improve forecasts shares. This, turn, help climate ensure growth economy. results existing on use other methods computational intelligence modeling behavior participants has been out. article presents study Four Stock Exchange were chosen forecasting: Centrenergo (CEEN); Ukrtelecom (UTLM); Kriukivs’kyi Vahonobudivnyi Zavod PAT (KVBZ); Raiffeisen Bank Aval (BAVL). following models experimental study: long short-term memory LSTM; convolutional network CNN; model combining two CNN consisting variational mode decomposition algorithm long-term (VMD-LSTM); VMD-CNN-LSTM deep learning based (VMD) networks. Estimates forecast quality metrics calculated. It concluded gives minimum error enterprises. also advisable VMD-LSTM exchanges countries an