IC Colloquium : Programming with Probabilistic Graphical Models
Event details
| Date | 15.12.2014 |
| Hour | 16:15 › 17:30 |
| Location | |
| Category | Conferences - Seminars |
By : Martin Vechev - ETH Zurich
Video of his talk
Abstract :
The increased availability of massive codebases (“Big Code”) creates an exciting opportunity for new kinds of programming tools based on probabilistic models. Enabled by these models, tomorrow’s tools will provide probabilistically likely solutions to programming tasks that are difficult or impossible to solve with traditional techniques.
In this talk, I will present a new approach for building such tools based on structured prediction with graphical models, and in particular, conditional random fields. These are powerful machine learning techniques popular in computer vision -- by connecting these techniques to programs, our work enables new applications not previously possible.
As an example, I will discuss JSNice (http://jsnice.org), a system that automatically de-minifies JavaScript programs by predicting statistically likely variable names and types. Since its release few months ago, JSNice has become a popular tool in the JavaScript community and is regularly used by thousands of developers worldwide.
Bio :
Martin Vechev is a tenure-track assistant professor of Computer Science at ETH Zurich where he leads the Software Reliability Lab (http://www.srl.inf.ethz.ch/). Prior to ETH, he was a Research Staff Member at the IBM T.J. Watson Research Center in New York (2007 - 2011). He obtained his PhD from Cambridge University in 2008. His research are in program analysis, program synthesis, concurrency, and applications of machine learning.
More information
Video of his talk
Abstract :
The increased availability of massive codebases (“Big Code”) creates an exciting opportunity for new kinds of programming tools based on probabilistic models. Enabled by these models, tomorrow’s tools will provide probabilistically likely solutions to programming tasks that are difficult or impossible to solve with traditional techniques.
In this talk, I will present a new approach for building such tools based on structured prediction with graphical models, and in particular, conditional random fields. These are powerful machine learning techniques popular in computer vision -- by connecting these techniques to programs, our work enables new applications not previously possible.
As an example, I will discuss JSNice (http://jsnice.org), a system that automatically de-minifies JavaScript programs by predicting statistically likely variable names and types. Since its release few months ago, JSNice has become a popular tool in the JavaScript community and is regularly used by thousands of developers worldwide.
Bio :
Martin Vechev is a tenure-track assistant professor of Computer Science at ETH Zurich where he leads the Software Reliability Lab (http://www.srl.inf.ethz.ch/). Prior to ETH, he was a Research Staff Member at the IBM T.J. Watson Research Center in New York (2007 - 2011). He obtained his PhD from Cambridge University in 2008. His research are in program analysis, program synthesis, concurrency, and applications of machine learning.
More information
Practical information
- General public
- Free
- This event is internal
Contact
- Host : Viktor Kuncak