This repository contains two lectures and a 3 session workshop on introducing machine learning concepts in the advanced physical chemistry module at UoE.
Dr Antonia Mey -- [email protected]
| Units | Materials |
|---|---|
| Unit_01: Dimensionality Reduction |  |
| Unit_02: Clustering | |
| Unit_02: Classification | |
| Unit_03: Neural Networks and PyTorch | |
| Unit_03: Deep Learning for Solubility Classification | |
-
Install anaconda.
-
Create a new environment:
conda create -n ml_chem -
Activate the environment:
conda activate ml_chem -
Install mamba to make the installation of packages faster.
conda install -c conda-forge mamba -
Install all the required packages with mamba:
mamba install -c conda-forge mdanalysis mdanalysistests mdanalysisdata nglview scikit-learn ipywidgets=7.6.0
Release: 13/02/2023
Report Deadline: 10/03/2023
Weight: 20%
- What is machine learning?
- Examples of machine learning (in Chemistry)
- Introduction to unsupervised learning:
- Clustering (k-means and others)
- How does actual input data look like?
- Molecular fingerprints and nomenclature Introduction to supervised learning:
- What is a classification problem?
- Unsupervised learning continued:
- Dimensionality reduction (PCA)
- t-SNE
- Regressions
- Classifications in practice:
- Random Forests
- Multilayer perceptrons
| Session | Materials |
|---|---|
| Dimensionality Reduction | |
| Clustering and Regressions | |
- Understand the main pillars of machine learning
- Know about different clustering techniques as part of unsupervised learning
- Be able to use common nomenclature used in machine learning
- Use Principle component analysis to reduce the dimensions of a data set
- Understand how a regression problem can be cast as a machine learning problem
- Be aware of how random forests and multilayer perceptrons can be used in a classification problem
A handout with additional resources can be found here.