JavaScript doesn’t tend to be people’s first choice for precise, performant numerical computing. However, it is the native language of the web and has a huge developer community. Because of this community and trends in on-device and Edge Computing, we’re seeing more machine learning libraries becoming available in JavaScript at the moment.
This workshop is an opportunity for ES6 JavaScript developers to delve into machine learning in a familiar language. It will provide a hands-on introduction to machine learning covering data pre-processing and feature engineering, common techniques in supervised and unsupervised learning and deep learning using libraries such as ml.js, convnet.js and keras.js.
We’ll work through examples in each area using Jupyter Notebooks running a Node.js kernel. We’ll also look at how to take these libraries into the browser and the potential for integrating them into web applications via visualisations.
We will provide a docker based environment running Jupyter for use during the workshop and to take away with you to continue development on your own.
Required audience experience: Intended audience are JavaScript developers with little exposure to Machine Learning or pairs with at least one JavaScript coder.
Objective of the training: The aim is to provide a practical introduction to ML in a familiar language. Participants should expect to take away new ML concepts, an idea of how to implement those in JS but also an idea of where to look in other languages / stacks for the same capabilities.
Keywords: JavaScript, Machine Learning, Deep Learning, keras.js, Jupyter, Node.js,