A carregar...
Publicação
Optimizing Credit Scoring Models in Face of Global Economic Uncertainty: A Comprehensive Risk Analysis in Banking loans
| Nome: | Descrição: | Tamanho: | Formato: | |
|---|---|---|---|---|
| 2.73 MB | Adobe PDF |
Autores
Orientador(es)
Resumo(s)
In the contemporary financial landscape, influenced by various global events, banks have
become increasingly cautious in extending credit, both for housing and for other consumer
purposes. In an increasingly uncertain economic world, banks must develop ever more robust
and effective models. This topic is of significant importance to explore and update, as banks
directly impact a substantial portion of the global population through housing and consumer
credit. This thesis aims to contribute to the development of models with current data, and
consequently, assist the population. In this research, Python programming language will be
employed to utilise a Machine Learning Approach for credit scoring analysis. Methods to be
utilised include: Logistic Regression; Random Forest; Gradient Boosting; XGBoost. To
determine the best model, the evaluation will be performed on four metrics: Accuracy; AUC
Score; Type I Error; Type II Error. The XGBoost method was the best performer on all evaluated
metrics. In the course of reviewing the selected literature, previous work were found that
explored this subject solely with the objective of identifying the best Machine Learning
method to create the optimal model for determining customer defaults. However, several
questions emerged: ‘Will machine learning models be better than Logistic Regression?’ ‘Is the
model that accepts more credit the safest for the bank, considering the Profit / Risk ratio?’
This thesis aims to answer these questions and determine not only the best model for the
bank, but also its profits.
Descrição
Dissertation presented as the partial requirement for obtaining a Master's degree in Statistics and Information Management, specialization in Risk Analysis and Management
Palavras-chave
Analytical Models Banking Machine Learning Credit Risk Credit Scoring Loan Defaults SDG 3 - Good health and well-being SDG 8 - Decent work and economic growth