Feras Saad
|
|
I am an Assistant Professor of Computer Science at Carnegie Mellon University. My research interests span programming languages, statistics, and artificial intelligence; where I develop new techniques for building scalable computing systems that enable flexible probabilistic modeling and sound probabilistic inference. This research has been applied to solve challenging problems in a variety of domains and is made available through multiple open-source software systems: see research and software for an overview.
Prior to CMU, I was a Visiting Research Scientist at Google. I received the PhD, MEng, and SB degrees in Electrical Engineering and Computer Science from MIT. My theses on probabilistic programming were recognized with the George M. Sprowls PhD Thesis Award in Artificial Intelligence + Decision Making (2023) and Charles & Jennifer Johnson MEng Thesis Award in Computer Science (2017).
Research Opportunities
I am actively recruiting self-motivated students and postdocs at all levels.- Prospective Students: If you are interested in working with me as a graduate student, please apply to the CMU CS PhD program and mention my name in your application.
- Current CMU Students: If you are already a graduate, masters, or undergraduate student at CMU, please send me an email and we can find a time to chat.
- Postdocs: Please send me an email with your CV, two representative papers, and a short description of your background and interests.
Probabilistic Programming.
Probabilistic programs that make random choices give us a uniquely
expressive approach to modeling and inference over complex
data-generating processes. My work develops new programmable systems that
aim to formalize, automate, and scale-up the hard aspects of probabilistic
modeling and inference within this paradigm.
[
Automated Probabilistic Model Discovery.
A central challenge in artificial intelligence is automating the process of
discovering accurate and interpretable probabilistic models that let us learn
and reason about data. We attack this problem from the lens of probabilistic program synthesis,
formalized as Bayesian structure learning over rich generative model families.
[
Statistical Estimators & Tests.
The traditional approach to analyzing statistical models is based on
hand-derived calculations that work with analytic mathematical expressions
of the underlying probability distributions.
However, as probabilistic programs become a more widespread representation
for complex models and inference algorithms, scalable simulation-based
techniques are needed to rigorously analyze their properties
through the black-box computational interfaces they expose.
[
Random Sampling Algorithms.
This thread explores fundamental computational limits of
random variate generation---a key operation that enables probabilistic
computation---including new algorithms that are theoretically optimal/near-optimal
(in entropy and error) and extremely efficient in practice.
[
Software + Applications.
A central motivation of my work is to develop performant, freely available