ISL Colloquium: Communication in Machine Learning

Topic: 
Communication in Machine Learning
Thursday, January 25, 2018 - 4:15pm
Venue: 
Packard 101
Speaker: 
John Cioffi (Stanford)
Abstract / Description: 

This Information Systems Seminar talk investigates the converse use of communication technologies and methods, sometimes in alternative forms or with different names, in machine learning. The next generation of internet communication has many uses for machine learning. This talk instead looks in more detail at some structures used in machine learning and draws analogies to methods previously used in communication. For instance, the recasting of a neural network with ReLu (rectifier linear unit) as having a state (linear or nonlinear) allows some analogy with hidden Markov models and state machines. Further, some of the back-propagation learning methods have analogies with forward-backward decoding algorithms that are in use as communication decoders. The question is thus posed as to if some of these communication methods might help certain applications of machine learning that are not viewed initially as communication problems. Some of these topics will be further examined in EE392AA (spring quarter), which can be used for EE MS Communications Depth sequence.


 

ISL Colloquium: The Information Systems Laboratory Colloquium (ISLC) is typically held in Packard 101 every Thursday at 4:15 pm during the academic year. Refreshments are usually served after the talk.

The Colloquium is organized by graduate students Martin Zhang, Farzan Farnia, Reza Takapoui, and Zhengyuan Zhou.

Bio:

John M. Cioffi - BSEE, 1978, Illinois; PhDEE, 1984, Stanford; Bell Laboratories, 1978-1984; IBM Research, 1984-1986; EE Prof., Stanford, 1986-present, now emeritus. Cioffi founded Amati Com. Corp in 1991 (purchased by TI in 1997) and was officer/director from 1991-1997. He currently also is on the Board of Directors of ASSIA (Chairman and CEO), Alto Beam, Tinoq, Collinear, and the Marconi Foundation, while serving several advisory boards. Cioffi's specific interests are in high-performance digital transmission. Cioffi's recognition includes: IEEE AG Bell, Kirchmayer, and Millenium Medals (2010, 2014, and 2000); Internet Hall of Fame (2014); Economist Magazine 2010 Innovations Award; International Marconi Fellow (2006); Member, US National and UK Royal Academies of Engineering (2001, 2009); IEEE Kobayashi and Armstrong Awards (2001 and 2013); BBWF Lifetime Achievement (2014), IEEE Fellow (1996); IEE JJ Tomson Medal (2000); 1991 and 2007 IEEE Comm. Mag. best paper; and numerous Conference Best-Paper awards. Cioffi has published over 600 papers and holds over 100 patents, of which many are heavily licensed including key necessary patents for the international standards in ADSL, VDSL, vectored VDSL, G.fast, DSM, LTE, Massive-MIMO, and various Wi-Fi methodologies.