Machine Learning in Production

Getting started

Start with creating Github and Kaggle accounts if you don't have them yet.

We'll work in Jupyter Notebooks and use a code editor. Make sure you have an editor set up with linting. If you're unsure what that means, install PyCharm Community Edition.

Get this repository to your account and the code on your machine: fork this repository to your account and clone your repository to your machine.

All the Python packages needed for this project are specified in a virtual environment. Make sure conda is installed, cd into the project root and create a new virtual environment:

$ conda env create -f environment.yml

This will create a virtual environment called ml-production and install the Python packages in it.

Now activate the virtual environment so that we can use it (you might have to use activate instead of source activate if you're on Windows):

(ml-production) $ source activate ml-production

We have some custom code for you in the folder shelter. Install the shelter package located in the project root with development mode:

(ml-production) $ python setup.py develop

Now start the Jupyter Notebook server so that we can start our Machine Learning:

(ml-production) $ jupyter notebook

Morning

We'll start with building a Machine Learning model. Open the Notebook 01-machine-learning-model.ipynb in the folder notebooks/ and follow the instructions.

Afternoon

Having made our first steps with building, we'll now focus on same development best practices. The material can be found in the Notebook 02-machine-learning-in-production.ipynb.