Paper: arXiv 2403.03785

Authors: Rambod Rahmani, Marco Parola, Mario G. C. A. Cimino

Abstract

Due to the recent increase in interest in Financial Technology (FinTech), applications like credit default prediction (CDP) are gaining significant industrial and academic attention. In this regard, CDP plays a crucial role in assessing the creditworthiness of individuals and businesses, enabling lenders to make informed decisions regarding loan approvals and risk management. In this paper, we propose a workflow-based approach to improve CDP, which refers to the task of assessing the probability that a borrower will default on his or her credit obligations. The workflow consists of multiple steps, each designed to leverage the strengths of different techniques featured in machine learning pipelines and, thus best solve the CDP task. We employ a comprehensive and systematic approach starting with data preprocessing using Weight of Evidence encoding, a technique that ensures in a single-shot data scaling by removing outliers, handling missing values, and making data uniform for models working with different data types. Next, we train several families of learning models, introducing ensemble techniques to build more robust models and hyperparameter optimization via multi-objective genetic algorithms to consider both predictive accuracy and financial aspects. Our research aims at contributing to the FinTech industry in providing a tool to move toward more accurate and reliable credit risk assessment, benefiting both lenders and borrowers.

Complexity vs Empirical Score

  • Math Complexity: 3.5/10
  • Empirical Rigor: 6.5/10
  • Quadrant: Street Traders — practical and empirical, lighter on theory

Why this score: The paper provides a practical, structured machine learning workflow for credit default prediction, combining Weight of Evidence encoding with ensemble methods and multi-objective genetic algorithms. While the mathematical formulations are fairly standard, the empirical focus on balancing predictive accuracy with financial metrics offers actionable, real-world utility.

Research Flowchart

  flowchart TD
    A["Research Goal: Improve Credit Default Prediction (CDP) & Financial Risk Assessment"] --> B["Data Preprocessing: Weight of Evidence (WoE) Encoding"]
    B --> C["Model Training: Train Multiple ML Model Families"]
    C --> D["Ensemble Learning: Combine Models for Robustness"]
    D --> E["Optimization: Multi-Objective Genetic Algorithms (Accuracy & Financial Aspects)"]
    E --> F["Outcomes: Accurate, Reliable Credit Risk Assessment Tool for FinTech"]

Cited References (33)

  1. Stock price forecasting over adaptive timescale using supervised learning and receptive fields (2018)
  2. A comprehensive review on NSGA-II for multi-objective combinatorial optimization problems (2021)
  3. An experimental comparison of classification algorithms for imbalanced credit scoring data sets (2012)
  4. Statistical and machine learning models in credit scoring: A systematic literature survey (2020)
  5. Credit scoring and its applications (2017)
  6. The value of big data for credit scoring: Enhancing financial inclusion using mobile phone data and social network analytics (2019)
  7. A necessary condition for a good binning algorithm in credit scoring (2014)
  8. Development and application of consumer credit scoring models using profit-based classification measures (2014) — link
and 25 more
  1. Measuring classifier performance: a coherent alternative to the area under the ROC curve (2009)
  2. Development and application of consumer credit scoring models using profit-based classification measures (2014)
  3. GitHub OptBinning repository, https://github.com/guillermo-navas-palencia/optbinning (2020) — link
  4. GitHub credit scoring code repository, https://github.com/MarcoParola/credit_scoring_benchmark (2023) — link
  5. Optimal binning: mathematical programming formulation (2020) — link
  6. Machine learning and decision support system on credit scoring (2020)
  7. An interval-valued approach to business process simulation based on genetic algorithms and the BPMN (2014)
  8. Weight-of-evidence through shrinkage and spline binning for interpretable nonlinear classification (2022)
  9. Approaches for credit scorecard calibration: An empirical analysis (2017)
  10. Calibrating deep neural networks using focal loss (2020)
  11. Machine learning with big data: Challenges and approaches (2017)
  12. Optuna: A next-generation hyperparameter optimization framework (2019)
  13. A survey of evolutionary algorithms for multi-objective optimization problems with irregular pareto fronts (2021)
  14. A deep learning approach for credit scoring using credit default swaps (2017)
  15. Credit default prediction modeling: an application of support vector machine (2017)
  16. Statistical methods for fighting financial crimes (2010)
  17. A survey on machine learning and statistical techniques in bankruptcy prediction (2018)
  18. Liquidity constraints, home equity and residential mortgage losses (2020)
  19. Image-Based Screening of Oral Cancer via Deep Ensemble Architecture (2023) — link
  20. Effects of Environmental Conditions on Historic Buildings: Interpretable Versus Accurate Exploratory Data Analysis (2023) — link
  21. Forecasting spot electricity prices: Deep learning approaches and empirical comparison of traditional algorithms (2018)
  22. Convolutional Neural Networks for Structural Damage Localization on Digital Twins (2023)
  23. Deep Learning of Structural Changes in Historical Buildings: The Case Study of the Pisa Tower (2022) — link
  24. Structural Damage Localization via Deep Learning and IoT Enabled Digital Twin (2022) — link
  25. Stock Price Forecasting Over Adaptive Timescale Using Supervised Learning and Receptive Fields (2018)

0 of 33 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2518 linked papers.