• A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Dynamic Model for Credit Cards Transactions Repayment Forecasting

Student: Suchkova Svetlana

Faculty: Faculty of Computer Science

Educational Programme: Financial Technology and Data Analysis (Master)

Year of Graduation: 2019

This paper presents dynamic model prototypes of two levels of granularity which are aimed at predicting credit cards transactions repayment in Sberbank. The best model settings appear to be decision tree-based ones: random forest and gradient boosting. More aggregate versions of models show high predictive power with R-squared > 0.96, while more detailed versions have moderate share of explained variance with R-squared ≈ 0.69 tested on an out-of-time sample. Both versions are applicable for further use in the construction of credit cards portfolio repayment forecasting model. However, these models can be improved in the ways of enriching the data with clients’ features, especially the information on their deposits and current accounts, and applying an asymmetric loss function in order to acknowledge the difference in forecasting errors costs for future retail loans portfolio.

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