This project aims to predict whether a client will subscribe to a term deposit with Banco de Portugal using machine learning. By analyzing a dataset containing various socio-economic factors, we developed a binary classification model to identify the likelihood of subscription.
Build a predictive model to classify clients based on their subscription status. Perform exploratory data analysis to uncover trends and patterns. Address class imbalance in the dataset to improve model performance. Tools & Technologies Python, Pandas, Scikit-learn, Matplotlib Logistic Regression, Random Forest Confusion matrix, Precision, Recall, F1-score
#Insights Age and job type are significant predictors, with older clients and certain job roles showing higher subscription rates. The dataset had a significant class imbalance, requiring careful handling to ensure accurate predictions. The model achieved an overall accuracy of 91%, with better performance on the majority class.
notebooks/: Jupyter notebooks for data analysis and modeling data/: Dataset used in the project scripts/: Python scripts for model development results/: Output metrics and visualizations