Understanding Machine Learning is a comprehensive textbook that introduces machine learning and its algorithmic paradigms in a principled manner. It covers the fundamentals, explores advanced topics, and provides practical insights into the field.
Recommended for: This textbook is recommended for advanced undergraduates and beginning graduates in statistics, computer science, mathematics, and engineering. It is also suitable for professionals in various industries who want to learn about machine learning without needing a background in computer science or programming.
You will:
- Gain a solid understanding of the fundamentals underlying machine learning.
- Learn about mathematical derivations that transform principles into practical algorithms.
- Familiarize yourself with a wide range of state-of-the-art algorithms.
- Understand when machine learning is relevant.
- Discover the prerequisites for successful application of ML algorithms.
- Learn which algorithms to use for different tasks.
Detailed Overview
The book is divided into four parts:
- Foundations: It provides a formal learning model, discusses uniform convergence, bias-complexity trade-off, VC-dimension, non-uniform learnability, and runtime of learning.
- From Theory to Algorithms: This section covers linear predictors, boosting, model selection and validation, convex learning problems, regularization and stability, stochastic gradient descent, support vector machines, kernel methods, multiclass and ranking problems, decision trees, nearest neighbor, and neural networks.
- Additional Learning Models: It explores online learning, clustering, dimensionality reduction, generative models, and feature selection and generation.
- Advanced Theory: This part delves into advanced concepts such as Rademacher complexities, covering numbers, proof of the fundamental theorem of learning theory, multiclass learnability, compression bounds, and PAC-Bayes.
The book combines rigorous theory and practical methods of machine learning, making it a valuable resource for both understanding the field and applying it to real-world problems. It emphasizes conceptual understanding while providing examples and exercises to solidify knowledge. The content progressively builds from basic supervised learning algorithms to advanced techniques like deep learning, covering various tasks, algorithms, and concepts.
“Understanding Machine Learning” has received praise from Bernhard Schölkopf, a prominent researcher in the field. It equips readers with a solid foundation to engage in current machine learning research and applications. Whether you are a university student or a professional interested in this transformative technology, this book serves as an essential guidebook to explore the fascinating world of machine learning.
Citation
Shai Shalev-Shwartz, Shai Ben-David. (2014). Understanding machine learning: Theory, algorithms. Cambridge University Press. https://www.cambridge.org/il/universitypress/subjects/computer-science/pattern-recognition-and-machine-learning/understanding-machine-learning-theory-algorithms
Licensing
This ebook, Understanding Machine Learning, is published by Cambridge University Press. The document is publicly available on the authors’ website: https://www.cs.huji.ac.il/~shais/UnderstandingMachineLearning/. Users may download a copy of the ebook for personal use only. It is not permitted to distribute the ebook. The document is freely accessible for online reading on the authors’ website.
Providing a hyperlink to the original source is generally considered legal, as:
- The content is already published and publicly accessible on the original website.
- This page is not reproducing or republishing the full content but only providing a link to the original source.
- This page is not modifying or altering the content in any way.
Download
