Agile Machine Learning: Effective Machine Learning Inspired by the Agile Manifesto

by Eric Carter, Matthew Hurst

Build resilient applied machine learning teams that deliver better data products through adapting the guiding principles of the Agile Manifesto.

Bringing together talented people to create a great applied machine learning team is no small feat. With developers and data scientists both contributing expertise in their respective fields, communication alone can be a challenge. Agile Machine Learning teaches you how to deliver superior data products through agile processes and to learn, by example, how to organize and manage a fast-paced team challenged with solving novel data problems at scale, in a production environment.

The authors’ approach models the ground-breaking engineering principles described in the Agile Manifesto. The book provides further context, and contrasts the original principles with the requirements of systems that deliver a data product.

What You'll Learn

Effectively run a data engineering team that is metrics-focused, experiment-focused, and data-focused
Make sound implementation and model exploration decisions based on the data and the metrics
Know the importance of data wallowing: analyzing data in real time in a group setting
Recognize the value of always being able to measure your current state objectively
Understand data literacy, a key attribute of a reliable data engineer, from definitions to expectations

Who This Book Is For

Anyone who manages a machine learning team, or is responsible for creating production-ready inference components. Anyone responsible for data project workflow of sampling data; labeling, training, testing, improving, and maintaining models; and system and data metrics will also find this book useful. Readers should be familiar with software engineering and understand the basics of machine learning and working with data.

268 pages, Kindle Edition

Published August 21, 2019