March 13 (Mon) @ 10:00am: "Robust Information Sharing and Processing: Efficiency, Reliability, and Interpretability," Homa Esfahanizadeh, Postdoctoral Researcher, MIT
Emerging applications and technologies require the processing of massive amounts of data and intensive computations within a reasonable timeframe. To meet this ever-growing demand, it is important to develop robust solutions for data transmission and processing that are fast, reliable, secure, and easily deployable. In the area of robust data processing, this talk will present a new scheme that realizes efficient, deadline-aware, and interpretable distributed computation. This scheme relies on a layering mechanism that enables approximations to be released over time, making the system robust to early terminations due to the deadlines. Additionally, an optimized scheduling mechanism is introduced that efficiently exploits the system heterogeneity. In the area of robust data communication, this talk will present a coding scheme for low-overhead error-correction that offers high reliability and low decoding latency. This scheme is based on optimizing a family of Low-Density Parity-Check codes, and relies on effectively reducing the design search space without eliminating optimal options along with a low-complexity inspection technique to identify the best code in the reduced search space.
Homa Esfahanizadeh is a postdoctoral researcher at the Massachusetts Institute of Technology (MIT). Currently, she works at the MIT Research Laboratory of Electronics (RLE), and her focus is on proposing robust solutions for data communication, processing, and learning over large-scale distributed information systems. Homa received her Ph.D. degree in Signals and Systems from the University of California, Los Angeles (UCLA) in 2019. She received her M.Sc and B.Sc degrees in Electrical Engineering from the University of Tehran in 2015 and 2012, respectively. Her research interests include coding and information theory, signal processing, and machine learning. Homa received the Electrical and Computer Engineering Department fellowship from UCLA in 2015 for her academic achievements. In 2018 and 2021, she won the Memorable Paper Award at the Non-Volatile Memories Workshop (NVMW), in the area of devices, coding, and information theory. In 2018, she won the 2018-2019 Dissertation Year Fellowship (DYF) at UCLA. In 2022, her paper was elected as one of the two best short papers at the International Conference on Cloud Networking (Cloudnet).
Hosted by: ECE Department
Submitted by: Amy Donnelly <email@example.com>