Primary DataAlternative Credit Scoringcs-alt-001
Alternative Credit-Scoring Feature Set — MFS-Derived
Period: 2022–2024Updated: 2025-10-08CSV · JSON
Abstract
Engineered feature set of 84 behavioural variables derived from anonymised MFS transaction logs, suitable for training alternative credit-scoring models. Includes default-flag labels on a 12-month outcome window.
Variables
account_idtx_count_30dp2p_ratiomerchant_ratiobalance_volatilitydefault_12m
Sources
- CFSBR research partnership
- Anonymised MFS logs
Data preview
| account_id | tx_count_30d | p2p_ratio | merchant_ratio | balance_volatility | default_12m |
|---|---|---|---|---|---|
| A-00012 | 142 | 0.42 | 0.31 | 0.18 | 0 |
| A-00013 | 38 | 0.71 | 0.08 | 0.34 | 1 |
| A-00014 | 96 | 0.38 | 0.44 | 0.12 | 0 |
| A-00015 | 24 | 0.62 | 0.04 | 0.41 | 1 |
| A-00016 | 184 | 0.29 | 0.56 | 0.09 | 0 |
🔒 Showing first 5 rows · Unlock the full dataset by licensing this product.
Cite this dataset
Centre for Fintech & Strategic Business Research (CFSBR). (2025). Alternative Credit-Scoring Feature Set — MFS-Derived [Data set]. CFSBR Data Bank. https://doi.org/10.67226/cfsbr.cs.001