edgarsmdn / mlce_book

Hands-on material for a Machine Learning in Chemical Engineering course

Machine Learning in Chemical Engineering

This repo contains the building material for a JupyterBook which is intended to serve as a template/prototype for the hands-on part of a Machine Learning in Chemical Engineering (MLCE) course. This was a collective effort between the Process Systems Engineering group at the Otto von Guericke University / MPI Magdeburg and the Optimisation and Machine Learning for Process Systems Engineering group at Imperial College London to share experiences and material used in the respective MLCE courses offered in these institutions.

To look at the book 📚💻 go to this link!

Contents

The book aims at covering application case-studies in chemical engineering of

• Supervised learning
• Unsupervised learning
• Reinforcement learning
• Data-driven optimization
• Hybrid modelling

Do you want to contribute?

If you have nice tutorials in the areas mentioned above reflecting case-studies in chemical engineering, we encourage you to share it with the community! 💪 For practical reasons, it is better if you submit your pull-request including a link to a working Colab Notebook.

Did you notice an error/typo?

Let us know! Submit your issue here and we will fix it. We encourage you to contribute to this resource!

Citation

To cite this JupyterBook use

@book{sanchez_chanona_ganzer_2023,
    title = {Machine Learning in Chemical Engineering},
    author = {Sanchez Medina, Edgar Ivan and del Rio Chanona, Ehecatl Antonio and Ganzer, Caroline},
    year = {2023},
    publisher = {JupyterBook},
    url = {https://edgarsmdn.github.io/MLCE_book/},
    DOI = {10.5281/zenodo.7986905}
}

or perhaps the more conventional:

• Sanchez Medina, Edgar Ivan, del Rio Chanona, Ehecatl Antonio, & Ganzer, Caroline. (2023). Machine Learning in Chemical Engineering (v1.0.0). Zenodo. https://doi.org/10.5281/zenodo.7986905

If you find this material useful give it a star ⭐ so that it can, potentially, help more people

mlce_book's Issues

Bug in hybrid CSTR

add sep=';' when reading the .txt data for the colab
Thanks to Henry from Indonesia to catch this bug!

scaling after splitting

In notebook 4, the scaling is performed before the splitting. This should be in reverse order by using only the "train" for scaling

"A quick reminder" on ANNs for ex. 4?

Shall we add a recap of ANNs to the beginning of exercise 4?
Exercises 2 and 3 have intros to KNN and GLM.

use chemical descriptors

change the use of fingerprints for chemical descriptors in notebook 4