Computer Science Speaker Series Master Calendar

View Full Calendar

CS Compiler Seminar: Saikat Dutta , "Testing Machine Learning Libraries in the Presence of Randomness

Event Type
Seminar/Symposium
Sponsor
Josep Torrellas
Location
In person 2124 Siebel Center for Computer Science or online: https://illinois.zoom.us/j/89400978467?pwd=NjV3ZFQrQ1JidTNyS0ZNUEVOcEtpUT09
Virtual
wifi event
Date
Nov 7, 2022   4:30 pm  
Views
40
Originating Calendar
Computer Science Speakers Calendar

We look forward to seeing you in person or online on Monday, November 7, at 4:30pm. Join in person at 2124 Siebel Center for Computer Science, 201 N. Goodwin Ave.

 

Abstract: 

Machine Learning is rapidly revolutionizing the development of many modern-day systems. However, ensuring the reliability of Machine Learning-based systems is challenging due to 1) the presence of non-determinism in internal (e.g., stochastic algorithms) and external (e.g., execution environment) components and 2) the lack of accuracy specifications. Traditional software testing techniques, while widely used to improve software reliability, cannot tackle these challenges since they rely on an assumption of determinism and lack domain knowledge.

The goal of my research is to develop novel testing techniques and tools to make Machine
Learning-based systems are more reliable. In this talk, I will present some of my PhD research work that focuses on automatically detecting bugs in Machine Learning libraries and improving the quality of develop-written tests in such libraries.

 

Bio:

Saikat is a PhD Candidate at UIUC, currently advised by Prof. Sasa Misailovic.  Saikat's research interests are in Software Engineering, with a focus on Software Testing and its intersection with Machine Learning. Saikat has won multiple awards for his research, including the Facebook PhD Fellowship, the 3M Foundation Fellowship, and Mavis Future Faculty Fellowship.

link for robots only