UKAN Machine Learning summer school 2019

This repository contains the PDF notes and Jupyter notebooks from the 2019 edition of the UK Acoustics Network (UKAN) Machine Learning summer school.

The intention of this summer school was to provide a grounding on basic machine learning, and to link this to acoustics applications.

We decided to place an emphasis on supervised learning and more specifically regression models, but touched on unsupervised learning towards the last day.

The course took place as follows:

• Day 1 - Linear models for regression
• Day 2 - Nonlinear models for regression and regularisation / ridge regression, and a brief introduction to cross-validation.
• Day 3 - Kernel Machines with an emphasis on kernel ridge regression
• Day 4 - Unsupervised learning, with an emphasis on density estimation.
• Day 5 - Machine Learning Challenge - itentifying outlying observations ona wave propagation dataset.

I have fixed some of the errors that crept into the slides, but if you spot any, please let me know ;)