RCTF 2022 Winter School

Welcome to the repository of the Machine Learning lecture given at the winter school of the Réseau Français de Chimie Théorique (RCTF) held from 17-28 January 2022. This repository includes the slides of the lecture about "Introduction to Supervised Machine Learning for Chemistry" and a jupyter notebook written in Python for the practical tutorial section.

http://www.chimie-theorique.cnrs.fr/spip.php?article994&lang=fr

Main topics covered in this tutorial

• Linear regression models

  • Ordinary least-squares;
  • Gradient descent;
  • Ridge regression.

• Non-linear regression models

  • Polynomial feature map;
  • The kernel trick;
  • Kernel ridge regression.

Requirements

The hands-on tutorial was designed to run in a jupyter notebook environment with

• python3 (tested with version 3.8.1)

The following libraries are requested for the tutorial:

• Math and data processing libraries:

  • pandas
  • numpy
  • scikit-learn

• For visualization:

  • seaborn
  • matplotlib
  • plotly

• Specialized packages:

  • mendeleev

The Jupyter notebook tutorial can be also executed in Google Colab.

References

1. Liu, X., Meijer, G. & Pérez-Ríos, J., A data-driven approach to determine dipole moments of diatomic molecules. Phys. Chem. Chem. Phys., 22, 24191 (2020)
2. Rupp, M., Machine Learning for Quantum Mechanics in a Nutshell. Int. J. Quantum Chem., 115, 1058 (2015)