This book provides a thorough overview of the evolution of privacy-preserving machine learning schemes over the last ten years, after discussing the importance of privacy-preserving techniques. In response to the diversity of Internet services, data services based on machine learning are now available for various applications, including risk assessment and image recognition. In light of open access to datasets and not fully trusted environments, machine learning-based applications face enormous security and privacy risks. In turn, it presents studies conducted to address privacy issues and a series of proposed solutions for ensuring privacy protection in machine learning tasks involving multiple parties. In closing, the book reviews state-of-the-art privacy-preserving techniques and examines the security threats they face.
|
|
|
1 | (14) |
|
1.1 What Is Machine Learning? |
|
|
1 | (3) |
|
1.2 Why Machine Learning Needs Privacy-Preserving Manner |
|
|
4 | (3) |
|
|
|
7 | (2) |
|
|
|
9 | (6) |
|
|
|
11 | (4) |
|
2 Secure Cooperative Learning in Early Years |
|
|
15 | (16) |
|
2.1 An Overview of Neural Network |
|
|
15 | (2) |
|
2.2 Back-Propagation Learning |
|
|
17 | (3) |
|
2.3 Vertically Partitioned Training Dataset |
|
|
20 | (5) |
|
2.3.1 Privacy-Preserving Two-Party Training |
|
|
20 | (1) |
|
|
|
21 | (1) |
|
|
|
21 | (3) |
|
2.3.4 Analysis of Security and Accuracy Loss |
|
|
24 | (1) |
|
2.4 Arbitrarily Partitioned Training Dataset |
|
|
25 | (6) |
|
2.4.1 BGN Homomorphic Encryption |
|
|
26 | (1) |
|
|
|
26 | (1) |
|
|
|
27 | (3) |
|
|
|
30 | (1) |
|
3 Outsourced Computation for Learning |
|
|
31 | (16) |
|
3.1 Outsourced Computation |
|
|
31 | (1) |
|
3.2 Multi-key Privacy-Preserving Deep Learning |
|
|
32 | (15) |
|
|
|
32 | (3) |
|
3.2.2 Homomorphic Encryption with Double Decryption Mechanism |
|
|
35 | (2) |
|
|
|
37 | (2) |
|
|
|
39 | (4) |
|
|
|
43 | (2) |
|
|
|
45 | (2) |
|
4 Secure Distributed Learning |
|
|
47 | (10) |
|
4.1 Distributed Privacy-Preserving Deep Learning |
|
|
47 | (4) |
|
4.1.1 Distributed Selective SGD |
|
|
48 | (1) |
|
|
|
49 | (2) |
|
4.2 Secure Aggregation for Deep Learning |
|
|
51 | (6) |
|
|
|
52 | (1) |
|
4.2.2 Technical Intuition |
|
|
53 | (1) |
|
|
|
54 | (2) |
|
|
|
56 | (1) |
|
5 Learning with Differential Privacy |
|
|
57 | (8) |
|
|
|
57 | (3) |
|
|
|
57 | (1) |
|
|
|
58 | (2) |
|
5.2 Deep Learning with Differential Privacy |
|
|
60 | (2) |
|
5.2.1 Differentially Private SGD Algorithm |
|
|
60 | (1) |
|
|
|
61 | (1) |
|
5.3 Distributed Deep Learning with Differential Privacy |
|
|
62 | (3) |
|
|
|
62 | (1) |
|
5.3.2 Estimating Sensitivity |
|
|
63 | (1) |
|
|
|
64 | (1) |
|
6 Applications---Privacy-Preserving Image Processing |
|
|
65 | (10) |
|
6.1 Machine Learning Image Processing for Privacy Protection |
|
|
65 | (1) |
|
6.2 Feature Extraction Methods of Machine Learning Image Processing |
|
|
66 | (1) |
|
6.3 Main Models of Machine Learning Image Processing for Privacy Protection |
|
|
67 | (8) |
|
6.3.1 Privacy-Preserving Face Recognition |
|
|
68 | (2) |
|
6.3.2 Privacy-Preserving Object Recognition |
|
|
70 | (2) |
|
6.3.3 Privacy-Preserving Classification |
|
|
72 | (2) |
|
|
|
74 | (1) |
|
7 Threats in Open Environment |
|
|
75 | (12) |
|
7.1 Data Reconstruction Attack |
|
|
75 | (4) |
|
|
|
76 | (1) |
|
|
|
77 | (2) |
|
7.2 Membership Inference Attack |
|
|
79 | (4) |
|
|
|
79 | (1) |
|
|
|
80 | (3) |
|
7.3 Model Stealing Attack |
|
|
83 | (4) |
|
|
|
84 | (1) |
|
|
|
85 | (1) |
|
|
|
86 | (1) |
|
|
|
87 | |
Jin Li is currently a professor and the vice dean of the Institute of Artificial Intelligence and Blockchain, Guangzhou University. He received his B.S. (2002) and M.S. (2004) from Southwest University and Sun Yat-sen University, both in Mathematics. He got his Ph.D. degree in information security from Sun Yat-sen University at 2007. His research interests include design of secure protocols in artificial intelligence, cloud computing (secure cloud storage and outsourcing computation), and cryptographic protocols. He served as a senior research associate at Korea Advanced Institute of Technology (Korea) and Illinois Institute of Technology (USA) from 2008 to 2010, respectively. He has published more than 100 papers in international conferences and journals, including IEEE INFOCOM, IEEE TIFS, IEEE TPDS, IEEE TOC, and ESORICS, etc. His work has been cited more than 11000 times at Google Scholar and the H-Index is 40. He served as an associate editor for several international journals, including IEEE Transactions on Dependable and Secure Computing, Information Sciences. He also served as the program chairs in the committee for many international conferences such as CSS 2019, ICA3PP 2018, CSE 2017, IEEE EUC 2017, and ISICA 2015. He received several National Science Foundation of China (NSFC) Grants, including NSFC Outstanding Youth Foundation.
Ping Li was born in May 1985 in Baojing Country of Hunan Province. She received her Ph.D. in School of Mathematics at Sun Yat-Sen University in June 2016 (Supervisor Prof. Zheng-An Yao) and joined the Guangzhou University as a postdoctoral fellow from July 2016 to December 2018 (Co-Supervisor Prof. Jin Li. Currently, she works at South China Normal University (Youth Talent). Her research fields are applied cryptography, cloud computing security, and privacy-preserving machine learning. Her current research direction contains cryptographic technologies, storage security and computation security in cloudcomputing, machine learning in securely outsourced computation, etc. She has published or accepted 20 academic papers, including 14 SCI papers and two ESI highly cited papers. She is undertaking the Youth Project of National Natural Science Foundation of China.
Zheli Liu received the B.Sc. and M.Sc. degrees in computer science from Jilin University, China, in 2002 and 2005, respectively. He received the Ph.D. degree in computer application from Jilin University in 2009. After a postdoctoral fellowship in Nankai University, he joined the College of Cyber Science of Nankai University in 2011. Currently, he works at Nankai University as an associate professor. His current research interests include applied cryptography and data privacy protection.
Xiaofeng Chen received his B.S. and M.S. in Mathematics from Northwest University, China, in 1998 and 2000, respectively. He got his Ph.D. degree in Cryptography from Xidian University in 2003. Currently, he works at Xidian University as a professor. His research interests include applied cryptography and cloud computing security. He has published over 100 research papers in refereed international conferences and journals. His work has been cited more than 4000 times at Google Scholar. He is in the Editorial Board of IEEE Transactions on Dependable and Secure Computing (IEEE TDSC), Security and Communication Networks (SCN), and Computing and Informatics (CAI), etc. He has served as the program/general chair or program committee member in over 30 international conferences.
Tong Li received his B.S. and M.S. from Taiyuan University of Technology and Beijing University of Technology, in 2011 and 2014, respectively, both in Computer Science & Technology. He got his Ph.D. degree in information security from Nankai University at 2017. After a postdoctoral fellowship in Guangzhou University, he currently is an associate professor in Nankai University. His research interests include applied cryptography and data privacy protection in cloud computing.
Daugiau apie elektronines knygas