Loan Approval Prediction

Project Overview

This project predicts whether a loan application will be approved or rejected based on applicant information and financial data. The goal is to build a binary classification model that accurately predicts loan approval while handling imbalanced data.

---

Dataset

• Source: Kaggle – Loan Approval Prediction Dataset
• Number of Entries: 4269
• Features (13 columns):

---

Tools & Libraries

• Python 3.x
• Pandas – Data manipulation
• NumPy – Numerical computations
• Scikit-learn – Modeling and evaluation
• Imbalanced-learn – Handling imbalanced data (SMOTE)
• Seaborn & Matplotlib – Data visualization

---

Project Workflow

1. Data Collection

• Downloaded dataset from Kaggle.
• Loaded CSV into a Pandas DataFrame.

2. Exploratory Data Analysis (EDA)

• Checked dataset shape, column types, and summary statistics.
• Analyzed target variable distribution to identify imbalance.
• Visualized categorical features vs loan status.
• Created correlation heatmap for numeric features.
• Confirmed there were no missing values in this dataset.

3. Data Preprocessing

• Dropped loan_id (identifier).
• Encoded categorical features using one-hot encoding: education, self_employed.
• Converted target variable loan_status to binary (1 = Approved, 0 = Rejected).
• Scaled numeric features using StandardScaler:

4. Train-Test Split

• Split dataset into 80% training and 20% testing.
• Stratified split to maintain class distribution:

5. Handling Imbalanced Data

Checked class distribution: Approved vs Rejected. Applied SMOTE to oversample minority class:

6. Model Training

• Trained three models on the resampled dataset:
• Logistic Regression
• Decision Tree Classifier
• Random Forest Classifier

7. Model Evaluation

• Evaluated models on test set using:
• Accuracy
• Precision
• Recall
• F1-score (main metric due to imbalance)
• Confusion matrix visualization

8. Model Saving

• Saved scaler and trained models using joblib for future use: