Optimal Experimental Design (OED) ·

Optimal experimental design, machine learning, Fisher information

This project provides

• convenient interfaces for experiments, metrics to evaluate them, functions and statistical models
• a library for statistical models, experiments and metrics
• in particular the pi-experiment
• a pipeline for benchmarking and comparing different experiments for different scenarios
• jupyter notebooks to easily run and learn how to use this projects content

This project is closely related to the "Parameter Individual Optimal Experimental Design and Calibration of Parametric Models" paper (DOI:10.1109/ACCESS.2022.3216364) where the background and mathematical framework used here is explained in detail. It also contains the notebook and its html used for the simulations in the paper.

Installing

Run in your terminal

git clone https://github.com/nicolaipalm/oed.git
cd oed/

Activate your favourite virtual environment and run

pip install -r requirements.txt

In order to set up a conda environment (recommended when working in jupyter notebooks) type in your terminal

conda create -n oed
conda activate oed
conda install pip
pip install -r requirements.txt

Getting started

The easiest way in order to get familiar with this project is by checking out the notebooks. Navigate to the notebooks sub-directory and follow the instructions given in the README.md.

Alternatively you can run the aging_model_pipeline.py file in the ./pipeline/aging_models subdirectory your favourite IDE.

Versioning

TBA

Tests

TBA

Licensing

TBA