Название исследуемой задачи:	Repeated learning in recommender systems
Тип научной работы:	M1P
Авторы:	Прозорова Лилия, Крехов Николай
Научный руководитель:	кандидат физико-математических наук, Хританков Антон Сергеевич
Научный консультант(при наличии):	Веприков Андрей Сергеевич

Abstract

This paper addresses the issue of evaluating the quality of recommender systems in the long term, taking into account the evolution of consumers and product assortments. We consider the dynamical system of changes in consumers and products over time. The main purpose of the study is to identify the conditions under which degeneracies in audience, assortment, or transaction distribution occur in a given repeated machine learning system, and how such phenomena depend on the learning algorithms and recommendation models. Using the obtained results, we presented a model that is able to increase the metrics in the recommendation systems without degenerating the distributions on products and customers. We conduct a series of computational experiments on the synthetic datasets, the results of the experiments correspond to the theoretical predictions derived from the dynamical model.

Research publications

Presentations at conferences on the topic of research

Software modules developed as part of the study

1. A python package mylib with all implementation here.
2. A code with all experiment visualisation here. Can use colab.