All notable changes to Emukit will be documented in this file.
- Added simple GP model for examples
- Bayesian optimization with unknown constraints
- Removed dependency on libomp
- Max value entropy search acquisition function
- Multi point expected improvement acquisition function
- Moved model free designs to core
- Profet implementation
- Added citation info
- QRBF for uniform and Gaussian measures
- uncertainty sampling acquisition for bq
- Bayesian Monte Carlo
- Bugfixes
- Doc fixes
- Added support for inequality constraints
- Fabolas as an example
- Bugfixes
- Confirmed support for Python 3.7
- Removed dependency on GPyOpt
- Implemented generic IntegratedHyperParameterAcquisition
- Added notebooks validation automation
- Random baseline for benchmarking
- Implemented a range of discrete optimizers
- Uniform measure and mutual information acquisition for BQ
- Added sample_uniform method to parameter space and individual parameter types
- Improved unit test coverage
- Various fixes in code, comments and notebooks