The corresponding notebooks to the lecture focus on first learning how to use Scikit-Learn, a widely-used machine learning package in Python, and then illustrate how to use it with scientific data that requires pre-processing, using molecular property prediction as an example.
This is a standalone version of the machine learning tutorial from the ALCF AI Training Series.
There are two ways to run the notebooks.
Binder will build the enviornment for you and host it on a cloud-hosted instance. Just click:
The environment.yml
file provided with this README describes how to build the environment with anaconda.
Once you have anaconda installed, build the environment by calling:
conda env create --file environment.yml
from the command line. Once installed, follow the instructions Anaconda generates to activate the environment and then launch Jupyter:
jupyter lab