Computer Science,
Control and
Geoinformation Doctorate

Seminar on November 11, 2020

Title

Dependability and Security in Probabilistic Data Structures and Machine Learning.

Speaker

Prof. Pedro Reviriego, Universidad Carlos III de Madrid

When and Where

November 11, 2020, from 15:00 to 17:00
Online on Teams

Abstract

The slowdown of CMOS technology scaling has put architectures and algorithms as the focus for future performance gains in computing systems. A promising approach at the algorithmic level are Probabilistic Data Structures (PDS) that use the application tolerance for small deviations in the results to reduce the implementation complexity. At the same time, Machine Learning (ML) algorithms are being widely used to exploit the huge amount of data available to improve system performance in many application domains.

Dependability and Security are key requirements for the adoption of both Probabilistic Data Structures and Machine Learning in safety critical systems. In this series of seminars, we will first provide an introduction to both PDS and ML focusing on security and dependability to then present recent works on these topics with the following tentative schedule:

  1. Bloom Filters Dependability and Security (16-Nov)
  2. Count Min Sketch Dependability and Security (23-Nov)
  3. Hyperloglog Dependability and Security (30-Nov)
  4. Machine Learning Fault Tolerance (7-Dec)

Speaker’s Short Bio

Pedro Reviriego received the M.Sc. and Ph.D. degrees in telecommunications engineering from the Technical University of Madrid, Madrid, Spain, in 1994 and 1997, respectively. From 1997 to 2000, he was an Engineer with Teldat, Madrid, working on router implementation. In 2000, he joined Massana to work on the development of 1000BASE-T transceivers. From 2004 to 2007, he was a Distinguished Member of Technical Staff with the LSI Corporation, working on the development of Ethernet transceivers. From 2007 to 2018 he was with Nebrija University. He is currently with Universidad Carlos III de Madrid working on high speed packet processing and probabilistic data structures focusing on performance, dependability and security.