St Petersburg University experts help school students to develop an algorithm of predicting stock values based on the news stream
An expert from St Petersburg University, Assistant Researcher Daria Klishevich, has helped the participants of ‘Big Challenges’ science and technology programme to develop an analytical service with transparent recommendations for stock brokers.
Research and Technology Programme ‘Big Challenges’
The service establishes a connection between the stock market quotes and global events allowing to take an informed decision about selling or buying stock shares. The project has been performed by the team within the field of ‘Big Data, Artificial Intelligence, Financial Technologies and Machine Learning’ at the ‘Sirius’ educational centre.
The Graduate School of Management at St Petersburg University took part in the project together with such leading Russian universities as: the Moscow Institute of Physics and Technology; the Skolkovo Institute of Science and Technology (Skoltech); Novosibirsk State University; and many others. ‘The Graduate School of Management at St Petersburg University is in close cooperation with the ‘Sirius’ educational centre supported by VTB Bank. We are confident that this collaboration will be of maximal use to each side. The programme participants have a great opportunity to receive up-to-date knowledge from the leading experts of the business school and create modern products under their mentorship. In its turn, St Petersburg University allows for the Graduate School of Management to meet talented school students that can become its future applicants. Together we create knowledge, form future leaders and change the world for the better,’ noted Olga Dergunova, Deputy President and Chairman of VTB Bank Management Board and Director of the Graduate School of Management.
People use different strategies in taking investment decisions — they take expert advice and read specialised mass media and Telegram-channels. As for the banks, they offer ‘financial consultants’ that develop recommendations with the help of artificial intelligence. However, by far not every person has trust in such services since the principle of their operation is hidden from the user.
School students participating in the ‘Big Challenges’ have developed their own programme that analyses the news, predicts changes in the quoted prices and gives advice on buying or selling the shares. The service also sorts the news according to the topics and shows their mood allowing for the user to evaluate the balance between positive and negative events. The news stream is updated once a day.
The service has a transparent algorithm of giving recommendations. The user will be able to read the news and understand what has affected the market, why the share value has risen or dropped.
Aleksei Riabykh, Project Director, Managing Director at VTB Bank
‘Apart from the buy-or-sell decision, you can get an aggregated collection of news for any time period and see the review of the Russian economy: top news, key persons and well-known companies. There are no analogies of the service in the Russian language yet,’ notes Aleksei Riabykh, Project Director, Managing Director at VTB Bank.
The school students downloaded the data from the news websites for the past 10 years including RIA News, Interfax, Kommersant, RBC. Then, they analysed them with the help of artificial intelligence in order to find proper names in the text and make connections between them. Then, they built diagrams and made a relevant classification of texts. The students also taught the artificial intelligence to make predictions based on the collected data.
‘We work closely with machine learning and create many models and neuron networks. I would like to work in this area, so I got interested in the technical side of the project. My tasks included thematic modelling, clustering and transforming the texts into vectors — the categories with which mathematics works,’ says Elizaveta Pushkareva, a ten-grade school student from Yekaterinburg. She came with a project of the service for diagnosing stroke based on the brain computer tomography and making treatment suggestions to the doctor.
St Petersburg University has been collaborating with the ‘Sirius’ educational centre since 2017. The University employees are among the founders, expert and advisory board members of the educational centre.
The ‘Big Challenges’ research and technology educational programme took place in the ‘Sirius’ on 5–28 July, 2021. 425 school students from 64 regions took part in it, which was the maximal number of participants since the programme start.
School students of the 8th—10th grades performed 80 projects in 12 fields that fall within the priority areas for the research and technological development of Russia: from agricultural biotechnologies to artificial intelligence. School students received various research, market and industry-specific tasks from over 70 partners of the ‘Big Challenges’, among which are the leading Russian universities and R&D institutes, state corporations, large companies and public organisations. In future, the projects can be further elaborated and used in the real economy or for research.
Skilful experts helped the teams to ‘package’ and present the project at the defence. ‘Under the mentorship of the VTB colleagues, school students made a service for analysing the news that helps to understand the interconnection between the news stream and stock values of the Russian companies. A user-friendly service with a laptop and smart phone version is a serious project to develop only within three weeks. The Graduate School of Management at St Petersburg University together with school students formulated the service cost proposal, target audience and the demand it covers. It is always useful to talk to the Sirius students, since we learn what our potential applicants are interested in,’ says Daria Klishevich, an expert from the Centre for International Business Studies at the Graduate School of Management at St Petersburg University.
Now the team has to integrate the developed algorithms on one website, complete the visual part and increase the performance of the service, so that it can withstand higher loads. The school students demonstrated a ready-made broker service at the project defence.