HSMA: Machine Learning Notebooks is a free online collection of the notebooks from the machine learning module of the Health Service Modelling Associates (HSMA) programme, with thanks to Mike Allen and the SAMueL team.
Rather than a book, it is a set of guided code examples that talk through the how and why of each step, giving you a library of code snippets to use when building your own machine learning projects. It covers:
- logistic regression
- tree-based models for classification and regression, including decision trees, random forests and boosted trees such as XGBoost and LightGBM
- neural networks with tensorflow
- explainable AI, including PDP, ICE and SHAP plots
- synthetic data
- calibrating and optimising models, including imputation, feature engineering, feature selection and hyperparameters
The examples are written in Python, mostly using the scikit-learn package, and include health datasets such as stroke thrombolysis data. Links to the lecture videos and slides are included for every session of the module, including those without notebooks.
View the notebooks
To view the notebooks, click on the image above or go to https://machinelearning.hsma.co.uk.
Brand new to Python? Consider starting with the HSMA book of Python first: