Название: Managing Machine Learning Projects (MEAP v8)
Автор: Simon Thompson
Издательство: Manning Publications
Год: 2022
Страниц: 361
Язык: английский
Формат: pdf, epub
Размер: 10.2 MB

Guide Machine Learning projects from design to production with the techniques in this unique project management guide. No ML skills required!

In Managing Machine Learning Projects you’ll learn essential Machine Learning (ML) project management techniques, including:

Understanding an ML project’s requirements
Setting up the infrastructure for the project and resourcing a team
Working with clients and other stakeholders
Dealing with data resources and bringing them into the project for use
Handling the lifecycle of models in the project
Managing the application of ML algorithms
Evaluating the performance of algorithms and models
Making decisions about which models to adopt for delivery
Taking models through development and testing
Integrating models with production systems to create effective applications
Steps and behaviors for managing the ethical implications of ML technology

Managing Machine Learning Projects is an end-to-end guide for project managers who need to deliver machine learning applications on time and under budget. It gives you a unique set of tools, approaches, and processes designed to handle the unique requirements of machine learning project management—all proven in practice to deliver success in full-scale business environments. You’ll follow an in-depth case study of a Bike Shop developing their first machine learning application and see how to put each technique into practice. Throughout, the book gives strong consideration to the ethical issues of ML, including data privacy, and community impact. You’ll learn how to avoid and mitigate these issues and keep your machine learning project from being exposed to failure.

about the technology
Companies of all shapes, sizes, and industries are investing in Machine Learning (ML). Unfortunately, something like 85% of all ML projects fail. Managing machine learning projects requires adopting a different approach than you’d take with standard software projects. You’ll need to account for large and diverse data resources, evaluating and tracking multiple separate models, and handling the unforeseeable risk of poor performance. Never fear—this book lays out the unique practices you’ll need to ensure your projects succeed.

about the book
Managing Machine Learning Projects is a comprehensive guide to delivering successful Machine Learning projects from idea to production. The book is laid out as a series of fictionalized sprints that take you from pre-project requirements and proposal development all the way to deployment. You’ll discover battle-tested techniques for ensuring you have the appropriate data infrastructure, coordinating ML experiments, and measuring model performance. With this book as your guide, you’ll know how to bring a project to a successful conclusion, and how to use your lessons learned for future projects.

The book is aimed at people who are being asked to take on a Machine Learning project that’s going to need a team of people working on it for a few months at least. While there are lots and lots of resources explaining how to apply the technology and algorithms of ML, there seems to be much less support for people pulling everything together and actually delivering these projects. Given the value of ML as a technology, more and more people are likely to find themselves in this position. This book was written to try to bridge the gap and provide a point of reference to help new ML project managers.

As you will find out, running an ML project is hard. ML algorithms produce unpredictable results, and using them introduces risk into a project. ML projects must handle large and complex data resources and will produce a lot of models that all must be evaluated. Additionally, ML projects present ethical challenges, which require careful management of all these data resources and models throughout the system’s lifecycle. My book describes approaches to handling these challenges and a process that works to reduce the risks and maximize reproducibility and accountability.