- Added new Attentive Non-stationary Kernel - Fixed Neural Kernel Non-stationary Kernel - Updated SGPR's update method to support all kernel functions
1.1.8
Key Updates
- Add option to skip inducing variables update in SSGP
1.1.7
Key Updates
- Update SSGP to use assign method-based updates - Update SSGP to use original Z init method - Update SSGP init to handle ndim data
1.1.6
Key Updates
- Fix package version
1.1.5
Key Updates
- Update get_model_params to include variables to control the number and training of inducing points when using an SGP - Add optional seed param to synthetic data gen method - Add a method to sample unlabeled points in a polygon defined with vertices
1.1.4
Key Updates
- Update get_model_params to use an SGP for large datasets and optionally return the trained model - Convert TSP start/end idx to nodes