SMT Optimization

A multi-fidelity constrained Bayesian optimization toolkit

Key Features

The SMT Optimization package offers a collection of surrogate-based optimization frameworks. The following frameworks are available:

Framework	Inequality Constraints	Equality Constraints	Multi-fidelity	As seen in
SEGO	Yes	Yes	No	https://doi.org/10.1080/03052150211751
MFSEGO	Yes	Yes	Yes	https://doi.org/10.2514/6.2019-3236

Getting Started

Prerequisites

smt-optim requires the following Python package to be installed:

1. Numpy pip install numpy
2. SciPy pip install scipy
3. SMT pip install smt

Installation

1. Clone the repo

git clone https://github.com/SMTOrg/smt-optim.git

2. Install smt-optim to your Python environment. In the root directory, type:

pip install -e .

Usage

See usage examples in the examples/ directory.

Please cite us when using SMT Optimization

If you are using SMT Optimization in your work, please cite the following paper.

Oihan Cordelier, Youssef Diouane, Nathalie Bartoli and Eric Laurendeau. "Multi-Fidelity Constrained Bayesian Optimization with Application to Aircraft Wing Design," AIAA 2025-3474. AIAA AVIATION FORUM AND ASCEND 2025. July 2025.