Patrick Jaillet - Research
Please contact me directly if you are interested in learning more about one of these activities.
main interests these days:
- online and data-driven optimization problems
-
- dynamic and real-time optimization in networks
-
- applications: internet, transportation, energy, finance
-
research group:
current
- Yossiri Adulyasak , post-doc, SMART, Singapore, 2013-now
-
- Setareh Borjian, master student, MIT CEE, 2012-now
-
- Iain Dunning, doctoral student, MIT ORC, 2011-now
-
- Chong Yang Goh, doctoral student, MIT ORC, 2012-now
-
- Swati Gupta, doctoral student, MIT ORC, 2011-now
-
- Dawsen Huang, doctoral student, MIT EECS, 2011-now
-
- Nikita Korolko, doctoral student, MIT ORC, 2012-now
-
- Maokai Lin, doctoral student, MIT ORC, 2009-now
-
- Xin Lu, doctoral student, MIT ORC, 2009-now
-
- Nicolai Ludvigsen, urop, MIT EECS, 2012-now
-
- Vahideh Manshadi, post-doc, MIT EECS and ORC, 2011-now
-
- Andrew Mastin, doctoral student, MIT EECS, 2010-now
-
- Ali Oran, post-doc, SMART, Singapore, 2011-now
-
- Jin Hao Wan, M.Eng., MIT EECS and Math, 2011-now
-
recent
- Jacob Cates, master student, MIT ORC/Draper, 2009-11
-
- Brian Crimmel, master student, MIT ORC/Draper, 2010-12
-
- Shen Shen, master student, MIT EECS, 2011-12
-
- Christian Therkelsen, m.eng, MIT EECS, Spring 2011
-
- Rico Zenklusen, post-doc, MIT EECS and Math, 2011-12
-
research interns
- Antoine Legrain, Ecole Centrale de Paris, 1/2011-7/2011
-
- Thibault Lehouillier, ENSIMAG, 2/2011-8/2011
-
- Pierre Jeremie, Ecole Polytechnique Paris, 3/2011-6/2011
-
- Alexandre Hollocou, Ecole Polytechnique Paris, 4/2012-8/2012
-
current research projects:
Below are some specific topics that could each lead to the definition of several research topics of interests for either interns/urops, graduate students, and/or postdocs.
- 1. Online Traveling Salesman and Hamiltonian Path Problems: Concentrating on autonomous spatial exploration and information harvesting problems, this research considers online version of classical combinatorial optimization problems (TSP, Hamiltonian path) with (i) incomplete and uncertain input streams and (ii) time-sensitive objectives. Goal is to design and analyze rigorous algorithmic solution strategies for these canonical problems.
- 2. Online Resource Allocation Problems: Similar in spirit to the previous project, but motivated by applications from telecom/internet (sponsored search auctions and online auctions, load balancing for content delivery networks, distributed caching problems, on-demand video/movie requests). Analyis of online versions of classical bipartite matching problems, as well as of other related problems such as the matroid secretary problems as well as more general linear programming problems.
- 3. Real-time Paths Tracking/Predictions and On-Demand Route Guidance Under Uncertainty: This overall activity is to develop novel algorithms using real-time data (from many heterogeneous sources) in order to (i) track and predict paths in dynamic transportation networks, and (ii) provide on-demand route guidance under uncertainty. Based on a combination of optimization, data-fusion, machine learning, and novel behavioral techniques. The aim is to develop rigorous data-driven algorithms and methodologies with provable properties, and practical implementations. Overview and current list of sub-projects with collaborators from Singapore.
- 4. Port Optimization: Our overarching goal is to develop data-centric models and algorithms for real-time port optimization. Our work is to include the development and application of a number of different methodologies, including data fusion and mining, model formulation and validation, on-line algorithmic development, and assessment of solution quality and impacts. We plan to apply these methodologies in the context of import and export-heavy and/or transshipment operations. An example research project in which we are interested involves using historical and current information to allocate containers to locations and subsequently plan container movements in order to minimize the number of unproductive moves. By predicting operational bottlenecks and providing tools for fully automated systems, we envision our approaches to be used by operators for real-time decision making.
current funding:
- Singapore MIT Research Alliance (SMART): Future Urban Mobility: Real-time Paths Tracking/Predictions and On-Demand Toute Guidance under Uncertainty (2010-15)
- Air Force Office of Scientific Research (AFOSR): Data-Drive Online and Real-Time Combinatorial Optimization (2010-13)
- National Science Foundation (NSF): Online Optimization for Dynamic Resource Allocation Problems (2010-13)
- Office of Naval Research (ONR): Online and Dynamic Optimization Problems under Uncertainty (2011-14)
- Office of Naval Research (ONR): Decentralized Online Optimization in Multi-Agent Systems in Dynamic and Uncertain Environments (2012-17)
[ Home |
General |
Research |
Teaching ]
Last modified March 2013.