Work in progress: educational implementations of machine learning algorithms.
All algorithms implemented in python with minimal dependencies. Efficient enough to use on real data, but with readability prioritized over speed.
All code should be readable if you have the following basic skills:
- A grasp of basic python. We'll stay away from the more advanced features, but things like simple lambda functions or list comprehensions shouldn't be beyond you.
- A grasp of basic numpy. We won't use any advanced features, but a good graph of basic things like broadcasting will be important. We will provide some notebooks to help you.
- An understanding of the underlying method. We won't explain everything from scratch, only how to get from an I-followed-the-lecture level understanding to an implementation. We will start with the algorithms explained in the mlvu.github.io lectures, so those lectures should give you a solid background to understand the algorithms.
Some modules have scripts that can be called from the command line. For instance.