Method
Credit scoring
Default risk scoring with reject inference and information value.
12 cases from 5 companies apply Credit scoring. 6 report results in production.
Typical tasks: Learn user and item embeddings, Optimize decisions, Evaluate models, Decide who gets an offer and 1 more. Companies: Alfa-Bank and 4 more. Often combined with: Logistic regression, Blending, Feature encoding, Gradient boosting, Population Stability Index and 20 more.
How One Product Helps Build Scoring for All of a Bank's Products (A DS Perspective)![]() | ![]() | banking | 2025-05 |
A Data-Driven Approach to Subscriber Scoring in Telecom![]() | ![]() | telecom | 2025-05 |
| Scoring on Non-Bank Data | — | fintech | 2024-05 |
| Reject Inference Methods: Credit Scoring at the Microfinance Company Dozarplaty | fintech | 2024-05 | |
| DL in finance: key results, production and new challenges | banking | 2022-06 | |
| Neural Network Models on Card Transactions | banking | 2022-06 | |
| Beating distribution shift in neural credit scoring | banking | 2022-06 | |
| Unified Neural Network Credit Scoring Model | banking | 2022-06 | |
| Neural network models on transaction data of a settlement account | banking | 2022-06 | |
| Neural network models on OFD receipt sequences | banking | 2022-06 | |
| Social Scoring and a Phone-Number-Based Approval Model | banking | 2022-06 | |
| Neural network approach for processing sequential data in a bank | banking | 2022-06 |

