Tensor Techniques in Signal Processing and Machine Learning: Advances and Opportunities

ABSTRACT

Many applications in signal processing and machine learning involve handling big data with multi-dimensional measurements that can be represented as tensors. Tensor decompositions are powerful tools for uncovering the low-dimensional structure in high-dimensional signals. This talk will begin with an introduction to classical tensor decomposition methods, then proceed to explore advanced tensor techniques applied in neural networks. Additionally, it will discuss recent progress in merging tensor-based signal processing with deep learning, emphasizing the opportunities arising from the combination of model-driven and data-driven approaches in machine learning.

BIOGRAPHY

Lin Chen received the Ph.D. degree in information and communication engineering from Shanghai Jiao Tong University, Shanghai, China in 2022. From 2023 to 2024, he was a Postdoctoral Fellow at Shanghai Jiao Tong University. From 2024 to 2025, he was a Postdoctoral Fellow with the Department of Electrical and Computer Engineering, Stevens Institute of Technology, Hoboken, NJ, USA, where he is currently a Research Scientist. His research interests include high-dimensional signal processing and machine learning, with applications in image processing and wireless communication.