Feras Saad

Doctoral Candidate
Computer Science and Artificial Intelligence Laboratory (CSAIL)
Department of Electrical Engineering and Computer Science (EECS)
Massachusetts Institute of Technology (MIT)

fsaad [at] mit.edu    Blog    Github

I am a second year PhD student (MEng/SB 2016) in Computer Science and Artificial Intelligence. I am a member of the Probabilistic Computing Project, advised by Vikash Mansinghka. My academic research is centered around the design and implementation of probabilistic computing systems, as well as algorithmic approaches to statistical inference in noisy, sparse, and multivariate data. From an applied perspective, I am interested in deploying these methods to public interest data.


My primary research projects, developed with colleagues at probcomp, include:
  • BayesDB, a probabilistic programming platform for probabilistic data analysis built on sqlite.
  • cgpm, a library of composable probabilistic models, which serves as the modeling and inference backed for BayesDB.
  • iventure, an interactive, jupyter-based front-end for BayesDB.
Some introductory tutorials in iventure summarizing basic use cases of BayesDB can be found below:
  • Exploratory analysis on Gapminder, a dataset of global macroeconomic indicators of education, poverty, environment and health.
  • Predictive analysis on a table of Earth satellites from the Union of Concerned Scientists.
For access, please refer to the probabilistic programming stack.
Sublime Text users: check out my productivity plugin AddRemoveFolder.

Publications and Preprints

A Bayesian Nonparametric Method for Clustering, Imputation, and Forecasting in Multivariate Time Series
Saad, F.A. and Mansinghka, V.K. arXiv preprint, arXiv:1611.07051, 2017. [Abstract, Paper, Supplement]

Time Series Structure Discovery via Probabilistic Program Synthesis
Schaechtle*, U.; Saad*, F.; Radul, A.; and Mansinghka, V. arXiv preprint, arXiv:1611.07051, 2017. [Abstract, Paper]

Probabilistic Search for Structured Data via Probabilistic Programming and Nonparametric Bayes
Saad, F.; Casarsa, L.; and Mansinghka, V. arXiv preprint, arXiv:1704.01087, 2017. [Abstract, Paper, Supplement]

Detecting Dependencies in Sparse, Multivariate Databases Using Probabilistic Programming and Nonparametric Bayes
Saad, F. and Mansinghka, V. Artificial Intelligence and Statistics (AISTATS), 2017. [Abstract, Paper, Supplement]

A Probabilistic Programming Approach To Probabilistic Data Analysis
Saad, F. and Mansinghka, V. Advances In Neural Information Processing Systems (NIPS), 2016. [Abstract, Paper]

Probabilistic Data Analysis with Probabilistic Programming
Saad, F. and Mansinghka, V. In review at the Journal of Machine Research, preprint at arXiv:1608.05347 (2016). [Abstract, Paper]


Charles & Jennifer Johnson Computer Science Master of Engineering Thesis Award, MIT EECS 2017.


Summer 2017: Instructor at the Probabilistic Programming for Advanced Machine Learning Summer School in Washington, DC.
Fall 2016: TA for 9.S915, Introduction to Probabilistic Programming, a graduate seminar at MIT.
Summer 2015: Instructor at the Probabilistic Programming for Advanced Machine Learning Summer School in Portland, Oregon.


I periodically keep notes and ideas in Random Seed.
During undergrad I wrote various op-eds for The Tech, MIT's dwindling student newspaper.
In the distant past, I was a sports journalist at Goal.com, the world's #1 football website.