Today, machine learning is being applied to a growing variety of problems in a bewildering variety of domains. A fundamental challenge when using machine learning is connecting the abstract mathematics of a machine learning technique to a concrete, real world problem. This book tackles this challenge through model-based machine learning which focuses on understanding the assumptions encoded in a machine learning system and their corresponding impact on the behaviour of the system.
The key ideas of model-based machine learning are introduced through a series of case studies involving real-world applications. Case studies play a central role because it is only in the context of applications that it makes sense to discuss modelling assumptions. Each chapter introduces one case study and works through step-by-step to solve it using a model-based approach. The aim is not just to explain machine learning methods, but also showcase how to create, debug, and evolve them to solve a problem.
Features:
- Explores the assumptions being made by machine learning systems and the effect these assumptions have when the system is applied to concrete problems.
- Explains machine learning concepts as they arise in real-world case studies.
- Shows how to diagnose, understand and address problems with machine learning systems.
- Full source code available, allowing models and results to be reproduced and explored.
- Includes optional deep-dive sections with more mathematical details on inference algorithms for the interested reader.