Loan Default Classifier - Build & Evaluate Models in Python

Predict Loan Defaults in Python - Real LendingClub Data, Full EDA, and ML Models

This Python notebook guides you through a full classification project using LendingClub loan data (2007–2010) to predict borrower repayment outcomes.

🔍 What’s Inside:

• Python notebook, underlying csv file and a Readme file with setup instructions.
• Full exploratory data analysis (EDA) with visualizations
• Outlier detection and label correction for clearer model interpretation
• Class imbalance handling (84/16 split)
• Model building with Decision Tree and Random Forest
• Precision-focused evaluation to minimize false positives
• Feature importance analysis and champion model selection

📊 Use Cases:

• Credit risk modeling for fintech and banking
• ML portfolio projects for job interviews
• Hands-on practice with classification metrics (Precision, Recall, F1)

📁 Requirements:

• Python 3.7+
• Jupyter Notebook
• Libraries: pandas, numpy, matplotlib, seaborn, sklearn

📎 Dataset Source: Public LendingClub data via Kaggle: Loan Data

🔎 Ideal for learners, analysts, and fintech professionals building real-world ML portfolios.

📘 Bonus: Includes a README setup guide - no extra installs needed. Also includes plain-English comments throughout the notebook to clarify outputs and support practical interpretation.