Welcome to the repository for the unsupervised learning tutorial lecture of the Machine Learning and Quantum Computing for Quantum Molecular Dynamics - MLQCDyn school held in September 2022 in the Université Gustave Eiffel, Paris. This repository contains a Jupyter Notebook with Python code to demonstrate, in practice, the application of the popular unsupervised learning methods (dimensionality reduction and clustering) to analyze molecular dynamics data and extract useful chemical insights. The popular MD17 database generated from ab initio molecular dynamics simulations is used as a motivating example for the analyses presented in the tutorial.
-
Dimensionality reduction
- Principal Component Analysis (PCA);
- Kernel PCA.
-
Clustering analysis
- Hard partition clustering with K-Means;
- Evaluation metrics for clustering.
The tutorial was designed to run in a jupyter notebook environment with
- python3 (tested with version 3.8.1)
The following Python packages are necessary to run the tutorial:
-
Math and data processing libraries:
- pandas
- numpy
- scikit-learn
-
For visualization:
- seaborn
- matplotlib
-
Specialized packages for chemistry:
The tutorial can be also executed on-line via Google Colab.
- T. Hastie, R. Tibshirani, and J. Friedman, The Elements of Statistical Learning: Data Mining, Inference, and Prediction, 2nd ed., Springer-Verlag, New York, 2009
- A. Glielmo, B. E. Husic, A. Rodriguez, C. Clementi, F. Noé, and A. Laio, Chem. Rev., 2021, 121 (16), 9722-9758
- M. Ceriotti, J. Chem. Phys., 2019, 150, 150901
- A. Wolf, and K. N. Kirschner, J. Mol. Model., 2013, 19, 539–549