6.3.6. VUB Credit Scoring

6.3.6.1. Summary

This dataset holds loans granted by a Romanian non-banking financial institution, published in anonymised form by the VUB Data Analytics Laboratory alongside Petrides et al. [1]. The goal is to predict whether a borrower will fall 45 or more days behind on a payment.

It has since become one of the standard benchmarks for instance-dependent cost-sensitive credit scoring [2] [3]. The dataset is bundled with Empulse and works offline.

Classes	2
Defaulters	3206
Non-defaulters	15711
Samples	18917
Features	16

6.3.6.2. Using the Dataset

The dataset is loaded through the load_vub_credit_scoring function. This returns a Dataset object with the following attributes:

• data: the feature matrix

• target: the target vector

• cost_matrix: a CostMatrix with default values pre-filled

• instance_costs: a dict of per-instance cost drivers ('cl', 'fp_cost')

• feature_names: the feature names

• target_names: the target names

• DESCR: the full description of the dataset

import numpy as np
import pandas as pd
from empulse.datasets import load_vub_credit_scoring
from empulse.metrics import Metric, Cost

dataset = load_vub_credit_scoring(backend=pd)

# replace with your own model's predicted probabilities
y_score = np.random.default_rng(0).uniform(size=len(dataset.target))

cost = Metric(dataset.cost_matrix, Cost())
default_lgd = cost(dataset.target, y_score, **dataset.instance_costs)
higher_lgd = cost(dataset.target, y_score, loss_given_default=0.9, **dataset.instance_costs)

6.3.6.3. Cost Matrix

	Actual positive \(y_i = 1\)	Actual negative \(y_i = 0\)
Predicted positive \(\hat{y}_i = 1\)	tp_cost \(= 0\)	fp_cost \(= r_i + -\bar{r} \cdot \pi_0 + \bar{Cl} \cdot L_{gd} \cdot \pi_1\)
Predicted negative \(\hat{y}_i = 0\)	fn_cost \(= Cl_i \cdot L_{gd}\)	tn_cost \(= 0\)

with

• \(r_i\) : loss in profit by rejecting what would have been a good loan

• \(\bar{r}\) : average loss in profit by rejecting what would have been a good loan

• \(\pi_0\) : percentage of non-defaulters

• \(\pi_1\) : percentage of defaulters

• \(Cl_i\) : credit line of the borrower

• \(\bar{Cl}\) : average credit line

• \(L_{gd}\) : the fraction of the loan amount which is lost if the borrower defaults

This is the credit scoring cost matrix of Bahnsen et al. [4], with an interest rate of 4.79%, a cost of funds of 2.94%, a term of 24 months and a loss given default of 75%. The loss given default stays symbolic; the other parameters are baked into 'fp_cost'.

Petrides et al. [1] derived their own costs from the institution’s average return on investment and loss given default per business channel. That is not possible from the published data: every monetary column, including the loan amount and the Expected_loss and Expected_profit columns, is standardised to zero mean and unit variance. Following Vanderschueren et al. [2], the loan amount is shifted to be strictly positive, \(Cl_i = Loan\_amount_i - \min_j Loan\_amount_j + 10^{-9}\), and used as the credit line. The resulting costs are in arbitrary units: compare them relative to each other, not as money.

6.3.6.4. Data Description

Variable Name	Description	Type
v1 – v8	Anonymised application variables	categorical
has_fico	Whether the applicant has a FICO score	binary
business_channel	Business channel through which the loan was granted (1, 2 or 3)	categorical
fico_score	FICO score, standardised; 0 when missing	numeric
loan_amount	Loan amount, standardised	numeric
monthly_income	Monthly income, standardised	numeric
age	Age of the borrower, standardised	numeric
gearing_coefficient	Gearing coefficient, standardised	numeric
max_gearing_ratio	Maximum gearing ratio, standardised	numeric
default	Whether the borrower fell 45 or more days behind on a payment (1 = yes, 0 = no)	binary

6.3.6.5. References

[1] (1,2)

Petrides, G., Moldovan, D., Coenen, L., Guns, T., & Verbeke, W. (2020). Cost-sensitive learning for profit-driven credit scoring. Journal of the Operational Research Society. https://doi.org/10.1080/01605682.2020.1843975

[2] (1,2)

Vanderschueren, T., Verdonck, T., Baesens, B., & Verbeke, W. (2022). Predict-then-optimize or predict-and-optimize? An empirical evaluation of cost-sensitive learning strategies. Information Sciences, 594, 400–415.

[3]

Ballegeer, M., Bogaert, M., & Benoit, D. F. (2025). Evaluating the stability of model explanations in instance-dependent cost-sensitive credit scoring. European Journal of Operational Research, 326(2), 630–640.

[4]

A. Correa Bahnsen, D. Aouada, B. Ottersten, “Example-Dependent Cost-Sensitive Logistic Regression for Credit Scoring”, in Proceedings of the International Conference on Machine Learning and Applications, 2014.