BoFire: Bayesian Optimization Framework Intended for Real Experiments
Authors: Johannes P. Dürholt, Thomas S. Asche, Johanna Kleinekorte, Gabriel Mancino-Ball, Benjamin Schiller, Simon Sung, Julian Keupp, Aaron Osburg, Toby Boyne, Ruth Misener, Rosona Eldred, Chrysoula Kappatou, Robert M. Lee, Dominik Linzner, Wagner Steuer Costa, David Walz, Niklas Wulkow, Behrang Shafei
JMLR 2025 | Venue PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Theoretical | This paper discusses the differences between Bo Fire and other BO implementations and outlines ways that BO research needs to be adapted for real-world use in a chemistry setting. It introduces an open-source Python package and describes its features and design for facilitating experiments, but does not present new empirical results, data analysis, or experimental validations conducted by the authors within this paper. |
| Researcher Affiliation | Collaboration | Johannes P. D urholt1, Thomas S. Asche1, Johanna Kleinekorte1, Gabriel Mancino Ball1, Benjamin Schiller1, Simon Sung1, Julian Keupp2, Aaron Osburg3, Toby Boyne4, Ruth Misener4, Rosona Eldred5, Chrysoula Kappatou5, Robert M. Lee5, Dominik Linzner5, Wagner Steuer Costa5, David Walz5, Niklas Wulkow5, Behrang Shafei5 EMAIL 1Evonik Operations Gmb H, DE, 2Boehringer Ingelheim Pharma Gmb H & Co. KG, DE, 3Heidelberg University, DE, 4Imperial College London, UK, 5BASF SE, DE. The authors are affiliated with both industrial companies (Evonik Operations GmbH, Boehringer Ingelheim Pharma GmbH & Co. KG, BASF SE) and academic institutions (Heidelberg University, Imperial College London). |
| Pseudocode | No | The paper does not contain any structured pseudocode or algorithm blocks. It describes the functionalities and architecture of the Bo Fire software framework in narrative text. |
| Open Source Code | Yes | Our open-source Python package Bo Fire combines Bayesian Optimization (BO) with other design of experiments (Do E) strategies... we have developed (and continue developing) the opensource software package Bayesian Optimization Framework Intended for Real Experiments or Bo Fire1. 1. https://github.com/experimental-design/bofire |
| Open Datasets | No | The paper introduces a software framework for experimental design but does not use or provide access information for any specific open datasets for its own evaluation or analysis within the paper. The examples provided are illustrative. |
| Dataset Splits | No | The paper does not present any empirical studies or use specific datasets for analysis, therefore, no dataset split information is provided. |
| Hardware Specification | No | The paper describes a software framework and its application in industrial chemistry but does not provide specific details about the hardware used to develop or run the software described within the paper. |
| Software Dependencies | No | The paper mentions several software packages and libraries such as Bo Torch (Balandat et al., 2020), Pydantic (Colvin, 2024), and Fast API (Ramírez, 2024), but it only provides publication years for associated papers or release dates, not specific version numbers for the software dependencies themselves. |
| Experiment Setup | No | The paper describes a software framework for experimental design and optimization, but it does not detail any specific experiments conducted by the authors in the paper, nor does it provide hyperparameters or training configurations for such experiments. |