daniel roy // droy

research

My research interests lie at the intersection of computer science and probability theory; I define, study and develop probabilistic programming languages to develop computational perspectives on fundamental ideas in probability theory and statistics. I am particularly interested in the use of recursion to define nonparametric distributions on structured objects like data structures; representation theorems that connect computability and probabilistic structures; and the complexity of inference.

NIPS*2008 workshop

I co-organized (with Vikash Mansinghka, John Winn, David McAllester and Josh Tenenbaum) a workshop on probabilistic programming at NIPS*2008.

seminars, working groups

Along with David Sontag, I run the MIT Machine Learning Tea seminar. If you would like to give a talk, please contact us. This term I am also running a Machine Learing in Church working group where we are systematically encoding interesting probabilistic models into Church to gain insight into probabilistic programming languages, improve our understanding of existing models, and identify interesting new models. We are placing an emphasis on recursive stochastic processes, which often arise in language models and are, incidentally, not expressible in standard graphical model notation.

teaching

S2008 - 6.437 Inference and Information
with Polina Golland and Greg Wornell

F2007 - 6.867 Machine Learning
with Tommi Jaakkola

F2003 - 6.035 Computer Language Engineering
with Martin Rinard and Saman Amarasinghe

personal

Hello, my name is Dan(iel) and I am a graduate student in the EECS PhD program in computer science at the Massachusetts Institute of Technology Computer Science and Artificial Intelligence Laboratory (CSAIL). I collaborate with my advisor Leslie Kaelbling as well as members of Josh Tenenbaum's Computational Cognitive Science group on work at the intersection of cognitive science and machine learning. My other interests include scalable AI, algorithmic complexity, logical representations, grounding, coding and information theory, automated software synthesis, and programming language theory. Outside of academia, I enjoy making films, producing electronic music, playing volleyball, squash, tennis and basketball, skimboarding, reading non-fiction, skiing, biking around on my single speed, dancing salsa, studying cello and traveling.

am i hot?

You may have heard of website I started called AmIHot.com. After it was featured on Howard Stern's radio show in 2000, its popularity sky-rocketed. I took spring term of my sophomore year at MIT off to work on AmIHot.com full time and make it profitable. Before I sold AmIHot in 2004 (now HotOrNot.com), our web servers were serving up millions of page views, handled by 50,000 lines of code running on multiple database and web servers.

volleyball

I played "opposite" on the MIT Men's Varsity Volleyball Team. I was captain of the Cambridge University Men's Volleyball team during my year abroad at Cambridge during the 2001-2002 season, the best season in its history. We won both the English Volleyball Association (EVA) championship and the British Universities Sports Association (BUSA) championionship, a feat not achieved by any British university in the previous decade. The EUSA win was Cambridge University's third ever (and first for volleyball). This earned us a spot at the European University Championships in Athens, Greece, where came in 7th in Europe. I personally set the season record for most points in a season (kills, aces and blocks). In honor of our hard work, the entire starting team was awarded "Full Blues," a distinction reserved for Cambridge's top athletes.

Source: Jesus College Virtual Tour

study abroad

I spent my junior year abroad under the Cambridge-MIT Exchange program. CMI/CME is a great program and I highly recommend it. If you are interested in participating, I am more than willing to discuss my experience.

valid xhtml

Valid XHTML 1.1 re-validate
Valid CSS /styles/css1, /styles/css2

preprints

The Infinite Latent Events Model,
David Wingate, Noah D. Goodman, Daniel M. Roy, and Joshua B. Tenenbaum. To appear in Uncertainty in Artificial Intelligence (UAI), 2009.
Preprint available upon request.

Computable exchangeable sequences have computable de Finetti measures, with Cameron Freer.
To appear in Lecture Notes in Computer Science (LNCS) series: Proceedings of Computability in Europe (CiE 2009).
Preprint available upon request.

publications

Exact and Approximate Sampling by Systematic Stochastic Search, Vikash Mansinghka, Daniel M. Roy, Eric Jonas, and Joshua Tenenbaum. To appear at AISTATS 2009. [ PDF, bibtex ]

The Mondrian Process, with Yee Whye Teh.
In Neural Information Processing Systems (NIPS*2008) 21, 2009. [ PDF, bibtex ]

A stochastic programming perspective on nonparametric Bayes, Daniel M. Roy, Vikash Mansinghka, Noah Goodman, and Joshua Tenenbaum. Nonparametric Bayesian Workshop at ICML, Helsinki, Finland, 2008.

Church: a language for generative models, Noah Goodman, Vikash Mansinghka, Daniel M. Roy, Keith Bonawitz, and Joshua Tenenbaum. In Uncertainty in Artificial Intelligence (UAI), 2008. [ PDF, bibtex ]

Bayesian Agglomerative Clustering with Coalescents, Yee Whye Teh, Hal Daumé III, and Daniel M. Roy. In Neural Information Processing Systems (NIPS*2007) 20, 2008. [ PDF, bibtex ]

Discovering Syntactic Hierarchies, Virginia Savova, Daniel M. Roy, Lauren Schmidt, and Joshua B. Tenenbaum. In Cognitive Science, 2007. [ PDF, bibtex ]

AClass: An online algorithm for generative classification, Vikash K. Mansinghka, Daniel M. Roy, Ryan Rifkin, and Joshua B. Tenenbaum. In Proceedings of the 11th International Conference on Artificial Intelligence and Statistics, 2007. San Juan, Puerto Rico. [ Abstract, PDF, bibtex ]

Efficient Bayesian Task-level Transfer Learning, Daniel M. Roy and Leslie P. Kaelbling. In Proceedings of the International Joint Conference on Artificial Intelligience (IJCAI). Hyderabad, India. January, 2007. [ PDF, PS, bibtex ]

Learning Annotated Hierarchies from Relational Data, Daniel M. Roy, Charles Kemp, Vikash Mansinghka, and Joshua B. Tenenbaum. In Neural Information Processing Systems (NIPS*2006) 19, 2007. [ PDF, PS, bibtex ]

Clustered Naive Bayes, Daniel M. Roy. MEng thesis, Massachusetts Institute of Technology, 2006. [ PDF, PS, bibtex ]

Enhancing Server Availability and Security Through Failure-Oblivious Computing, Martin Rinard, Cristian Cadar, Daniel Dumitran, Daniel M. Roy, Tudor Leu and William S. Beebee, Jr.. In Proceedings of the 6th Symposium on Operating Systems Design and Implementation. San Francisco, CA. December, 2004. [ PDF, PS, bibtex ]

A Dynamic Technique for Eliminating Buffer Overflow Vulnerabilities (and Other Memory Errors), Martin Rinard, Cristian Cadar, Daniel Dumitran, Daniel M. Roy and Tudor Leu. In Proceedings of the 2004 Annual Computer Security Applications Conference. Tucson, AZ. December, 2004. [ PDF, PS, bibtex ]

Efficient Specification-Assisted Error Localization, Brian Demsky, Cristian Cadar, Daniel M. Roy and Martin C. Rinard. In Proceedings of the Second International Workshop on Dynamic Analysis. Edinburgh, Scotland. May, 2004. [ PDF, PS, bibtex ]

Efficient Specification-Assisted Error Localization and Correction, Brian Demsky, Cristian Cadar, Daniel M. Roy and Martin C. Rinard. MIT CSAIL Technical Report 927. November, 2003. [ PDF, PS, bibtex ]

Implementation of Constraint Systems for Useless Variable Elimination, Daniel M. Roy and Mitchell Wand. Research Science Institute. August, 1998. [ PDF, PS, bibtex ]

Professional Activities

IJCAI 2009 Program Committee, AISTATS 2009 Program Committee, NIPS 2008 Reviewer (Top Reviewer), ICML 2008 Program Committee, NESCAI Reviewer 2007, JMLR Reviewer 2006/07/08

view my curriculum vitæ

colleagues

Leslie P. Kaelbling, research advisor.

Martin C. Rinard, former advisor.

Daniela Rus, academic advisor.

Rahul Sarpeshkar, former academic advisor.

(coauthors) William Beebee, Keith Bonawitz, Cristian Cadar, Hal Daumé III, Brian Demsky, Daniel Dumitran, Cameron Freer, Noah Goodman, Eric Jonas, Leslie Kaelbling, Charles Kemp, Tudor Leu, Vikash Mansinghka, Ryan Rifkin, Martin Rinard, Virginia Savova, Lauren Schmidt, Yee Whye Teh, Josh Tenenbaum, David Wingate

academics

Doctoral Student
Electrical Engineering and Computer Science

summer '09 enrollment

6.THG

PhD Thesis

past enrollment

6.001	Structure of Computer Programs
6.002	Circuits and Electronics
6.003	Signals and Systems
6.004	Computational Structures
6.034	Artificial Intelligence
6.035	Compilers (espresso)
6.041	Probablistic Systems Analysis
6.050J	Information and Entropy
6.111	Digital Electronics
6.170	Lab in Software Engineering
6.253	Convex Analysis and Optimization
6.270	Autonomous Robot Competition
6.821	Programming Languages
6.825	Techniques of Artificial Intelligence
6.836	Embodied Intelligence
6.867	Machine Learning (TA)
6.868	Society of Mind
6.905	6.111 (extra units)
6.907	Compilers (extra units)
6.981	T.A. for 6.035, 6.867, 6.437
6.986	Inference and Information
8.01	Physics I - Mechanics
8.022	Physics II - E and M
9.520	Statistical Learning Theory and Applications
9.66	Computational Cognitive Science
17.01	Justice
18.022	Multi-variable Calculus
18.03	Differential Equations
18.062/6.042	Mathematics for Computer Science
18.100B	Analysis
18.125	Measure and Integration (OCW)
18.404/6.840	Theory of Computation
18.410/6.046	Algorithms
18.415/6.854	Advanced Algorithms
18.575	Model Theory
18.701	Algebra I
21L.009	Shakespeare
21M.301	Harmony and Counterpoint I
21M.302	Harmony and Counterpoint II
21M.361	Composing with Computers
21M.421	MIT Symphony (Stellar)
21M.775	Hip Hop
21W.785	Communication in Cyberspace
CMI.6.E1	Electronic Circuits ^
CMI.6.E2	Power Electronics ^
CMI.6.E3	Device Physics ^
CMI.6.E4	Feedback and Control ^
CMI.6.E5	Modulation, Networks and Optics ^
STS.035	History of Computing

* Requires MIT Security Certificate
^ Taken at Cambridge University, UK