GPdoemd: A Python package for design of experiments for model discrimination

Abstract

Model discrimination identifies a mathematical model that usefully explains and predicts a given system's behaviour. Researchers will often have several models, i.e. hypotheses, about an underlying system mechanism, but insufficient experimental data to discriminate between the models, i.e. discard inaccurate models. Given rival mathematical models and an initial experimental data set, optimal design of experiments suggests maximally informative experimental observations that maximise a design criterion weighted by prediction uncertainty. The model uncertainty requires gradients, which may not be readily available for black-box models. This paper (i) proposes a new design criterion using the Jensen-Rényi divergence, and (ii) develops a novel method replacing black-box models with Gaussian process surrogates. Using the surrogates, we marginalise out the model parameters with approximate inference. Results show these contributions working well for both classical and new test instances. We also (iii) introduce and discuss GPdoemd, the open-source implementation of the Gaussian process surrogate method.

Equations of State in Chemical Enginnering Using Python

“…First, we have to define the problem statement to solve by defining a function; then, we check it; finally, we issue a command along with correct syntax to solve it. [2] For example, we take the following equation and solve this with the help of python programming language There are various simple steps which we have to remember while solving the problem statement.…”

Pyomo.DOE: An open‐source package for model‐based design of experiments in Python

“…Unfortunately, the gPROMS MBDoE tool is currently limited to only parameter precision objectives. For Python users, there are a handful of open source MBDoE tools, including GPdoemd, 46 EFCOSS, 47 Python‐MBDOE, 48 and Pydex. 49 GPdoemd focuses on model discrimination, using data‐driven Gaussian process surrogate models, which helps circumvent some of the computational challenges associated with directly optimizing science‐based models. EFCOSS (Environment for Combining Optimization and Simulation Software) provides a Python interface for optimization, parameter estimation, and DoE using simulation models.…”

Simplified algorithms for adaptive experiment design in parameter estimation

“…More recently, publication of software has provided direct access to Bayesian design methods methods, potentially reducing the effort required for implementation. [5][6][7][8][9] The Bayesian optimal design method centers around locating maxima of the utility, U (d), which expresses the goals of the measurement as a function of candidate setting designs d. Several authors have identified computation of the utility as a particularly difficult part of Bayesian experiment design that prohibits * rmcmichael@nist.gov its use. [3,4,10,11] Importantly, for a sequential design to be preferable to a static design, the cost of implementing and running the design processes must not exceed the value of saved resources.…”