Machine Learning as a Microservice

Guest Speaker: Brenton Mallen. Brenton is an Ocean Engineer turned Data Scientist. He has spent the last three and a half years performing in-depth research and building both real-time and batch machine learning systems in the internet bot mitigation space.

Abstract:

This will be a walkthrough of building machine learning model and an associated microservice to execute a the model and return it's output. The model will be accessible both as a web app as well as a RESTful API. It will be using Python (specifically Flask and Zappa) along with the AWS serverless architecture (Lambda and API Gateway). The example case will be the classic Kaggle problem of determining if someone would have survived the Titanic (https://www.kaggle.com/c/titanic/data).