Managing Machine Learning Projects

Sometimes people manage machine learning projects as if they were software development projects. Sometimes they use approaches that have evolved and "just work for them." And sometimes they don't manage them at all! But the community "in the know" has come up with approaches and techniques that work to remove risk from ML projects and create high quality outcomes without burning out valuable teams in the process. I've written this all up in the book Managing Machine Learning Projects: From Design to Deployment (Manning, 2022). In this talk, I'll give an overview of the methodology and the motivations behind it. Attendees will learn: - How to get a project started; - What the team needs to focus on up front; - Where the risks come from; - How to handle the model development process; - Why evaluation is fundamental to success; - How and why to integrate ethics into the process; - What needs to be considered when integrating ML into applications.