This folder contains the source codes for ODBO algorithms
- bo.py: Naive Bayesian optimization to generate the next set of queries points
- featurization.py: Generate the feature vectors for different scenarios of protein datasets
- gp.py: Gauassian process regression model constructions, inclding the GP with GP likelihood and GP with studentT likelihood.
- initialization.py: Algorithm to find suitable initial set of measurements to be measured in experiments
- plot.py: Plotting functions to plot the confusion matrix for XGBOD accuracy and BO curves
- prescreening.py: The XGBOD search space prescreening algorithm
- regressions.py: Surrogate modeling with GP or RobustGP
- run_exp.py: Wrapped functions to pack the BO search of each iteration
- test.py: Test functions to make sure the package is installed correctly
- turbo.py: Trust region Bayesian optimization algorithm to generate the next set of queries points
- utils.py: Useful functions