Fathi Salem
American electrical engineer
From Wikipedia, the free encyclopedia
Fathi M. Salem is an American electrical engineer and academic who is a professor of electrical and computer engineering at Michigan State University, where he leads the Circuits, Systems and Artificial Neural Networks research group.[1] He is also affiliated with the MSU Neuroscience Program.[2] In 1996, he became an IEEE Life Member and Fellow "for contributions to the development of tools for the analysis and design of nonlinear and chaotic circuits and systems".[3]
University of California, Davis (MS)
Neural networks
Nonlinear circuits
Fathi M. Salem | |
|---|---|
| Alma mater | University of California, Berkeley (PhD) University of California, Davis (MS) |
| Known for | Recurrent neural networks Blind signal separation |
| Scientific career | |
| Fields | Electrical engineering Neural networks Nonlinear circuits |
| Institutions | Michigan State University |
Education
Salem received a PhD in electrical engineering and computer sciences from the University of California, Berkeley in 1983. He earned a Master of Science degree in electrical engineering from the University of California, Davis in 1979.[4]
Research
Salem's research focuses on neural networks and learning systems, blind signal deconvolution and extraction, dynamical systems and chaos, and integrated CMOS sensing and processing.[2] His work on blind source recovery established a state-space framework for the problem using Kullback–Leibler divergence as a performance functional.[5][6]
His more recent work has focused on recurrent neural networks, including developing simplified variants of long short-term memory (LSTM) architectures with reduced parameters.[7][8]