madoc-q/ghana-loan-default-prediction

This repository contains code for analyzing credit risk in Ghana and Germany using machine learning models. The code is based on a dataset of loan applications and includes preprocessing, model training, and evaluation scripts. # ghana-loan-default-prediction This repository contains code for analyzing credit risk in Ghana and Germany using machine learning models. The code is based on a dataset of loan applications and includes preprocessing, model training, and evaluation scripts. This repository contains code for analyzing credit risk datasets from both Germany and Ghana. The datasets used are: credit_risk_dataset.csv: German credit risk dataset, which includes 1,000 loan applications and associated data such as age, income, and credit history. loan_port.csv: Ghanaian credit risk dataset, which includes loan information for micro-lending institutions. Methodology The study focused on building and testing machine learning models to predict loan defaults. The models were evaluated using metrics such as accuracy, precision, recall, and F1 score. The code for analyzing these datasets is split into two notebooks: finalcode_german.ipynb: Notebook containing code for analyzing the German credit risk dataset. finalcode_ghana.ipynb: Notebook containing code for analyzing the Ghanaian credit risk dataset. The repository also includes a sample loan portfolio (loan_port.csv) for use in the analysis. Usage To use this repository, clone it to your local machine and open the desired notebook (finalcode_german.ipynb or finalcode_ghana.ipynb) in a Jupyter environment. Make sure to have the necessary dependencies installed before running the code. github.com