AIMS CDT - Michaelmas Term 2020, Week 4
This repository contains material related to the practical sessions and assessment of the Machine Learning module within the Autonomous Intelligent Machines and Systems (AIMS) CDT taught course at the University of Oxford.
For more information please visit the course web page.
These are hands-on tutorial sessions that are grouped into four parts reflecting the structure of the module covering eight lectures over four days. See under each practical directory for more information.
- L1-L2: linear regression
- L3-L4: differentiable programming
- L5-L6: image classification
- L7-L8: probabilistic programming
In order to follow the practicals you need to have the following installed.
- A Python runtime. Anaconda with Python 3.8 is highly recommended.
- Jupyter, SciPy, Numpy, Matplotlib -- These are standard modules included in Anaconda by default.
- PyTorch and Pyro -- These can be installed by following instructions in their websites.
Alternatively you can use a Google Colab notebook, within which PyTorch and the other dependencies are available by default and you can install Pyro by running !pip install pyro-ppl in a regular code cell. Colab notebooks run the code in the cloud and do not require any installation in your local machine.
The demonstrators for the practicals are Benjamin Moseley and Lewis Smith.
The course is assessed by a take-home assignment that needs to be submitted. See the assessed assignment directory for the instructions. Deadline: 9 November 2020