Information Theoretic Learning-based Filter
Description
This book provides a comprehensive and in-depth exploration of adaptive filtering algorithms based on the Information Theoretic Learning (ITL). As a powerful alternative to traditional second-order statistical methods, ITL-based adaptive filtering algorithms are particularly effective in dealing with non-Gaussian noise. The book systematically introduces core ITL criteria such as minimum error entropy and maximum correntropy and extends these principles to the field of multidimensional signal processing and nonlinear adaptive filtering, demonstrating their effectiveness through modeling real-world signals like wind speed and temperature. In addition to single-node filtering, this book thoroughly investigates distributed adaptive filtering, addressing collaborative learning across networked systems. It further integrates graph signal processing, allowing for efficient modeling and analysis of signals defined on irregular or structured domains. Together, these contributions showcase ITL as a unified and powerful learning framework, advancing adaptive filtering theory and methodology across linear, nonlinear, distributed, and graph-based signal processing environments.
More details
Persons
Haiquan Zhao (Senior Member, IEEE) received the B.S. degree in applied mathematics and the M.S. and Ph.D. degrees in signal and information processing from Southwest Jiaotong University, Chengdu, China, in 1998, 2005, and 2011, respectively. Since 2012, he has been a Professor with the School of Electrical Engineering, Southwest Jiaotong University. From 2015 to 2016, he was a Visiting Scholar with the University of Florida, Gainesville, FL, USA. He is the author or co-author of more than 280 international journal papers (SCI indexed) and owns 56 invention patents. His current research interests include information theoretical learning, adaptive filters, adaptive networks, active noise control, Kalman filters, machine learning, and artificial intelligence. Dr. Zhao has won several provincial and ministerial awards and many best paper awards at international conferences or IEEE Transactions. He has served as an Active Reviewer for several IEEE Transactions, IET series, signal processing, and other international journals. He is currently a Handling Editor of Signal Processing, and also an Associate Editor for IEEE Transaction on Audio, Speech and Language Processing, IEEE Transactions on Systems, Man and Cybernetics: System, IEEE Signal Processing Letters, IEEE Sensors Journal, and IEEE Open Journal of Signal Processing.
Xinyan Hou was born in Gansu, China, in 1995. He received the Ph.D. degree in electrical engineering at Southwest Jiaotong University, Chengdu, China. His research interests include adaptive signal processing and distributed estimation.
Xiaoqiang Long was born in Gansu, China, in 1995. He received the Ph.D. degree with the School of Electrical Engineering, Southwest Jiaotong University, Chengdu, China. His research focuses on adaptive signal processing.
Content
Introduction.- Maximum Correntropy Criterion .- Decorrelation Recursive Maximum Correntropy Criterion.- Combinatoric Correntropy Criterion.- Fast Minimum Error Entropy with Fiducial Points.- Asymmetric Correntropy Criterion.- Hybrid Geometric Algebra Correntropy Criterion.- Kernel Generalized Maximum Correntropy with Variable Centers.- Diffusion Kernel Generalized Maximum Correntropy with Variable Centers.- Nonlinear Graph Diffusion Maximum Mixture Correntropy.- Conclusion.