- Theory of Probability and Statistics (basic)
- Need to have some basic knowledge about probability and statistics, such as normal distribution, Laplace distribution, combinations, permutations, etc.
- Machine Learning (medium)
- Need to know the basic machine learning concepts, such as supervised learning (e.g., classification), unsupervised learning (e.g., regression, clustering), and machine learning techniques, such as Support Vector Machine (SVM), Logistic Regression, Linear Regression, K-means, and neural networks, etc.
- Python (medium)
- Need to know the basic syntax of Python and how to write and debug Python code. The reader also needs to be familiar with certain scientific computation and machine learning packages, such as NumPy, Scikit-learn, PyTorch, TensorFlow, etc.
- Java (medium)
- Need to know the basic syntax of Java and how to write and debug Java code.
- The reader will learn different privacy-preserving machine learning techniques, such as secure multiparty computation (MPC), compressive privacy, differential privacy (DP), local differential privacy (LDP), database security and privacy, etc.
- The reader will learn how to implement and deploy different privacy-preserving machine learning techniques, such as differential private principal component analysis, locally differential private deep neural network, etc.
- The reader will also learn and get an understanding of how to design a tailor-made privacy-preserving machine learning algorithm and their own privacy-preserving machine learning algorithms and systems by reading the showcases and projects in this book.
-
Part 1
- Part 1 covers the basics of privacy-preserving machine learning and differential privacy. Chapter 1 discusses privacy considerations in machine learning with an emphasis on the dangers of private data being exposed. Chapter 2 introduces the core concepts of differential privacy along with some widely adopted differential privacy mechanisms that serve as building blocks in various privacy-preserving algorithms and applications. Chapter 3 covers the advanced design principles of differentially private machine learning algorithms and presents a case study.
-
Part 2
- Part 2 looks at another level of differential privacy called local differential privacy and at generating synthetic data to ensure privacy. Chapter 4 introduces the core concepts and definitions of local differential privacy. Chapter 5 looks at the more advanced mechanisms of local differential privacy, focusing on various data types and real-world applications, and then presents another case study. Chapter 6 focuses on generating synthetic data for machine learning tasks.
-
Part 3
- Part 3 covers the next-level core concepts required to build privacy-assured machine learning applications. Chapter 7 introduces the importance of privacy preservation in data mining applications, looking at privacy protection mechanisms widely used in data mining for processing and publishing data. Chapter 8 discusses widely used privacy models in data mining and their threats and vulnerabilities. Chapter 9 focuses on compressive privacy for machine learning, discussing its design and implementation. Finally, chapter 10 puts the concepts from all the previous chapters together to design a privacy-enhanced platform for protecting and sharing research data.