Machine Learning-Based Risk Assessment and Default Prediction in P2P Lending Platforms
Smart Human-centered Emerging Research in Machine Intelligence · 29.04.2026 · 2 (1) · 48-59
DOI: 10.68012/shem.v2i1.56 ↗Kurz gefasst
Fachartikel in der Zeitschrift Smart Human-centered Emerging Research in Machine Intelligence, online erschienen am 29.04.2026. Verfasst von Bhupesh Rawat und 5 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
This study investigates machine learning-based risk assessment and default prediction in peer-to-peer (P2P) lending platforms, addressing increasing concerns related to credit risk, information asymmetry, and platform sustainability in digital financial ecosystems. The primary objective is to evaluate the effectiveness of machine learning models in predicting default risk and to identify key borrower, loan, and platform-level determinants influencing default outcomes. To achieve this objective, a mixed-method approach is employed, integrating quantitative analysis of publicly available loan-level data with qualitative case studies of selected P2P platforms. The quantitative component utilizes advanced machine learning algorithms, including ensemble learning methods, to model and predict default probability, while the qualitative analysis explores platform practices in risk evaluation, pricing strategies, and default management mechanisms. The findings demonstrate that machine learning models significantly outperform traditional credit scoring methods in predicting default risk, particularly when incorporating alternative data sources and behavioral features. Key determinants of default include borrower income stability, debt-to-income ratio, loan tenure, interest rate, and platform-specific risk policies. Furthermore, qualitative insights reveal that transparent risk communication and proactive monitoring mechanisms are critical in reducing default rates and enhancing platform resilience. In conclusion, the study highlights that the integration of machine learning techniques into risk assessment frameworks enhances predictive accuracy and supports sustainable P2P lending operations. It also underscores the importance of model transparency, explainability, and regulatory alignment to strengthen trust among stakeholders in digital lending environments.
Zitieren
APA
Rawat, B., Nathalie, J., Manongga, D., Fransiso, G., Moana, F. P. O., & Sunarya, P. A. (2026). Machine Learning-Based Risk Assessment and Default Prediction in P2P Lending Platforms. Smart Human-centered Emerging Research in Machine Intelligence, 2(1), 48-59. https://doi.org/10.68012/shem.v2i1.56
Vancouver
Rawat B, Nathalie J, Manongga D, Fransiso G, Moana FPO, Sunarya PA. Machine Learning-Based Risk Assessment and Default Prediction in P2P Lending Platforms. Smart Human-centered Emerging Research in Machine Intelligence. 2026;2(1):48-59. doi:10.68012/shem.v2i1.56
BibTeX
@article{Rawat2026,
author = {Rawat, Bhupesh and Nathalie, Julia and Manongga, Danny and Fransiso, Gabriel and Moana, Fitra Putri Oganda and Sunarya, Po Abas},
title = {Machine Learning-Based Risk Assessment and Default Prediction in P2P Lending Platforms},
journal = {Smart Human-centered Emerging Research in Machine Intelligence},
year = {2026},
volume = {2},
number = {1},
pages = {48--59},
doi = {10.68012/shem.v2i1.56}
}
EUVIMED