← Back to Projects
Bayesian Optimization Lab
PythonBoTorchNumPySciPyJupyterMatplotlibSeabornGaussian Processes
A comprehensive implementation of Bayesian Optimization (BO) for hyperparameter optimization, developed as part of the EDAP30 Advanced Applied Machine learning course. The project demonstrates both theoretical understanding and practical application of BO using Gaussian Processes as surrogate models for optimizing black-box functions.
Key Features
- ✓Implementation of Bayesian Optimization using Gaussian Processes as surrogate models
- ✓Expected Improvement (EI) acquisition function for optimization guidance
- ✓Real-world application on wine quality dataset for hyperparameter optimization
- ✓Visualization of the optimization process and uncertainty estimates
- ✓Comprehensive theoretical and practical understanding of BO concepts
Challenges & Solutions
- ↳Balancing exploration vs. exploitation in the optimization process
- ↳Applying theoretical concepts to real-world hyperparameter optimization tasks
- ↳Handling high-dimensional search spaces efficiently
Project Screenshots