Replication Data for: Optimizing an expensive multi-objective building performance problem: Benchmarking model-based optimization algorithms against metaheuristics with and without surrogates.

This dataset contains all generated samples for a multi-objective optimization benchmark on a realistic building performance simulation problem. The samples are saved in JSON files. Every file contains the results of an independent optimization run.<br/><br/>
The JSON log files are organized into two folders. <em>AA_BPS_Benchmark</em> contains the logs from the benchmark conducted using the Building Performance Simulation software TRNSYS for function evaluations. <em>BB_Surrogate_Benchmark</em> contains the files created using regression models for function evaluation, and CC_BPS_Random_Samples contains the logs used to train the surrogate models (see the <a href="https://darus.uni-stuttgart.de/file.xhtml?fileId=321422">README.txt</a> file for more information).