FastBO

Code for paper Experimenting, Fast and Slow Bayesian Optimization of Long-term Outcomes with Online Experiments

Installation

To install the code clone the repo and install the dependencies as

cd FastBO
python3 -m pip install -r requirements.txt

Running Benchmarks

The benchmarks/ directory contains code for running the numerical experiments described in the paper. The benchmark problems are defined in problem_facotory.py and the file benchmark_spec.json contains the input config of running benchmarks. To run these experiments

python3 run_benchmark benchmark_spec.json

License

This code is MIT Licensed, as found in the LICENSE file.