The results of the Joint Project of SCILA and VTB were Published

The journal PeerJComputerScience (IF = 2.41) published an article with the results of a joint project of the SCILA and VTB Bank.

The paper describes the developed algorithm that uses news in the largest Russian media to predict changes in the exchange rate of stock securities on the Russian market. The STTM (Stock Tonal Topic Modeling) algorithm uses thematic modelling methods and keyword sentiment detection for news analysis, and also determines the dynamics of Russian companies' stock prices.

The co-author of the publication, SKILA Leading Researcher Sergey Koltsov, notes that the resulting model is more accurate than 26 other models that do not take into account news and market conditions.

The code of the developed algorithm is in the open access.