M01 Nano-Tera.ch | Data science techniques and machine learning for EDA

Printer-friendly versionPDF version
Location / Room: 


Pierre Vandergheynst, EPFL, CH (Contact Pierre Vandergheynst)

Download handouts here (Handouts are available for attendees only! The password has been sent to you by email or you may ask for the password at the on-site registration desk.)

Data science and machine learning have been instrumental in the development of new applications that affect everyone's life. At the same time, it opens new research challenges but also new opportunities for improving optimisation algorithms. As many other fields, EDA has also been impacted, with several recent works that have proposed to use machine learning towards effective design strategies.

This tutorial proposes to give an introduction to data science and machine learning, with an emphasis on their potential benefits in EDA. The tutorial will put a specific focus on graph-based machine learning techniques, and on learning and inference from time-series data, which are two recent branches of machine learning that have high potential in design and system optimisation.