Dynamic Recommendations Using Reinforcement Learning Methods

Student: Burmistrov Roman

Educational Programme: Financial Technology and Data Analysis (Master)

Year of Graduation: 2019

Services of personalized online recommendations are constantly striving to improve the quality of recommendations of their products, in order to increase the pool of users and their loyalty. The preferences of users are constantly changing and therefore traditional methods of recommendations face difficulties in dynamically changing environment. In particular, this problem occurs when recommending news. In this paper, the task of recommending a news category is presented in the form of a problem of a contextual multi-armed bandit. To improve the quality of the news category recommendation in this work we offer a model together with corresponding contextual features. The quality of the proposed model is demonstrated on the news data life.ru in comparison with the standard approaches for solving the problem of a multi-armed bandit.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses