Professor Mingyi Hong elevated to IEEE Fellow

Professor Mingyi Hong was recently elevated to Fellow of the IEEE (class of 2025). The honor recognizes his “contributions to optimization in signal processing, wireless communication and machine learning.”

Hong’s research focuses on contemporary issues in optimization, information processing and training, and alignment and application for foundation models such as language and diffusion models. His contributions have made a significant impact on the development of efficient optimization algorithms for highly complex problems that arise in signal processing. He has designed generic optimization algorithms that can be customized to specific applications. The Block Successive Upper-bound Minimization (BSUM) framework developed by Hong breaks down highly non-convex problems into a series of simple ones and is now adopted for solving many contemporary signal processing problems such as dictionary learning, tensor decomposition, and others. He has also been a pioneer in the use of deep neural networks (DNNs) to replace the use of algorithms for some wireless transceiver design problems at low computational costs. The paper documenting this work was recognized with an IEEE SPS Best Paper Award in 2022. It was titled, “Learning to optimize: training deep neural networks for interference management.” The approach demonstrated in the paper is now a broadly used tool for solving computationally intense wireless communication problems. 

In the current scenario where technology advances have enabled individualized and decentralized access to data, distributed algorithms have played a key role in their smooth operation. Hong’s popularization of massively parallelizable distributed algorithms has significantly contributed towards such optimal operation. His proposal of an inexact distributed optimization framework by improving a classical algorithm called Alternating Direction Method of Multipliers (ADMM) is now a benchmark algorithm for solving energy management problems that arise in smart grid control, and problems that arise in wireless communications, and large-scale machine learning. The paper documenting this work was recognized with an IEEE SPS Best Paper Award in 2021.

Previously, Hong was honored by the 2022 IEEE Signal Processing Society (SPS) with the Pierre-Simon Laplace Early Career Technical Achievement Award which honors individuals who have made significant technical contributions within the field early in their career. The award in particular recognizes Hong’s “contributions to non-convex, distributed and learning-based optimization for signal processing.”

He has been the recipient of the 2020 Facebook (now Meta) Research Award, the 2020 IBM Faculty Research Award, and a Best Paper award of the International Consortium of Chinese Mathematicians. He was also a best paper finalist for the Young Researchers in Continuous Optimization Program of the Mathematical Optimization Society in 2013 and 2016.  He is currently an Amazon Scholar.

Share