The notebook in this repo creates a number of different models using the well-known Titanic dataset. It builds models using both scikit-learn and tensorflow, and uses GPU acceleration if available.

It demonstrates the use of a number of different data visualization tools, including pandas-profiling and graphviz.

The repo has 3 additional files: ez.json, build/Dockerfile, and build/requirements.txt that instruct the ez CLI on how to provision a VM to run this repo. The Dockerfile builds on top of the tensorflow/tensorflow image on DockerHub, and installs a few additional packages described in requirements.txt.

If you have ez installed and configured, all you need to do to run the repo is:

ez env go -g https://github.com/jflam/data-science

If you want to be able to edit the files, you should first fork the repo into your GitHub account and use this command:

ez env go -g [email protected]:<your user id>/data-science