This project focuses on developing a Loan Approval Prediction Model that predicts whether a loan application will be approved based on various factors, including applicant income, loan amount, credit history, and more. The model has achieved an accuracy of 83%, demonstrating its effectiveness in predicting loan approvals.
- Python: Programming language
- Pandas: Data manipulation and analysis
- NumPy: Numerical computing
- Matplotlib: Data visualization
- Seaborn: Statistical data visualization
- scikit-learn: Machine learning library
- StandardScaler: For data scaling
- LabelEncoder: For converting categorical data to numeric values
In this phase, the dataset was examined to understand the underlying patterns and relationships between variables. Key statistics were gathered to inform data preprocessing steps.
- Handling Missing Values: Implemented imputation techniques to fill in missing data.
- Outlier Detection: Normalization was used to fix outliers and improve model performance.
- Data Scaling: Features were scaled to ensure uniformity and better accuracy during model training.
The model was trained using various algorithms from scikit-learn, including:
- Decision Trees
- Naive Bayes
The dataset was split into 80% for training and 20% for testing sets to evaluate the model's performance effectively.
Model performance was evaluated using accuracy metrics to ensure reliable predictions. The final model achieved an accuracy of 83% on the test dataset.