Bank Marketing Campaigns

Predicted whether bank clients would subscribe to a term deposit using 11,162 records and 17 features. Performed end-to-end ML: exploratory data analysis, outlier removal, 3 missing-value strategies (KNN Imputer, random sampling, dropping), label encoding, and MinMax scaling. Trained and compared 6 models: Naive Bayes, KNN, Decision Tree, Random Forest, SVM, and Logistic Regression. Best result: Random Forest at 84.9% accuracy with GridSearchCV hyperparameter tuning.

Tech: Python, Scikit-learn, Pandas, Seaborn, Plotly, Matplotlib