It's one thing to understand ML, it's another to be able to use ML, and it's something entirely different to be able to BUILD ML. This hands on workshop is designed for the latter purpose.

At the DSML group we love to engage with open source and encourage a mindset of ambitious building. We'll use this approach to guide our way to learning about common design principles in lots of ML codebases, and how they interact to create the different tools ML Engineers and Data Scientists use on a daily basis.

Topics covered will include:

- Different layers of abstraction commonly used in projects like scikit-learn, tensorflow, and others

- How mathematical principles are codified into ML algorithms

- When python is and is not used, and tool choices when speed is of paramount importance

- How unit testing and ops are typically handled during pull requests

This is meant to be very hands on, and people should expect to spend lots of time poking around through github repos and staring at lots of examples of live code. The goal of this workshop is to give peope a birds eye view of what it means to actively contribute to an in-production ML codebase.

‍