Top Cited Article in Informatics Engineering Research: October 2020
CV
1. Jason K. Johnson
1192B 41st Street
Los Alamos, NM 87544 USA
home: (505) 412-4162
office: (505) 665-7816
email : jasonj383@gmail.com
http://ssg.mit.edu/group/∼jasonj
http://cnls.lanl.gov/External/people/Jason Johnson.php
Research Interests
graphical models, network optimization, machine learning, statistical signal and image pro-
cessing, statistical physics, combinatorial optimization, multiscale methods, information
theory, convex optimization and analysis.
Education
Massachusetts Institute of Technology, Cambridge MA.
Ph.D. Electrical Engineering and Computer Science, 2008.
S.M. Electrical Engineering and Computer Science, 2003.
S.B. Physics, 1995.
Graduate Research Advisor: Prof. Alan Willsky
Undergraduate Thesis Advisor: Prof. Edward Farhi
Professional Experience
Postdoctoral Fellow/Research Associate 2008-Present
Dr. Michael Chertkov Los Alamos, NM
Center for Nonlinear Studies & Theoretical Division T-4, Los Alamos National Laboratory
Director-Funded Postdoctoral Fellow (2009-2011) independently-funded, competitive 2-year
appointment made by selection committee annually. I researched combinatorial, variational
and multiscale approaches to approximate inference in graphical models, learning planar
Ising models, optimization and control of power transmission networks. I was a co-organizer
of the 2009 Physics of Algorithms Workshop, Santa Fe NM (http:/cnls.lanl.gov/poa),
have co-advised two graduate summer interns and have contributed to several research grant
proposals.
Research Assistant 2000-2008
Prof. Alan Willsky Cambridge, MA
Laboratory for Information and Decision Systems, MIT
I researched tractable inference and learning methods for graphical models, with applica-
tions to large-scale estimation problems in remote sensing. I played a central role in devel-
oping walk-sum analysis of Gaussian inference algorithms (belief propagation and iterative
methods), the Lagrangian relaxation method and convergent iterative message-passing for
estimation in graphical models and the maximum-entropy relaxation method and relaxed
iterative scaling algorithm for graphical model selection.
2. Summer Internship Summer 2005
Dr. Evan Fortunato Burlington, MA
Alphatech, Inc.
Developed Lagrangian relaxation technique for hypothesis pruning in the multiple-hypothesis
testing approach to multi-target tracking.
Teaching Assistant Fall 2003
Prof. Tommi Jaakkola Cambridge, MA
Department of Electrical Engineering and Computer Science, MIT
Taught recitation sections and assisted in development of problem sets for introductory
machine learning course.
Member of Technical Staff 1995-2000
Dr. Robert Washburn Burlington, MA
Alphatech, Inc.
Algorithm development and prototyping. C/C++ programming. Automatic target recog-
nition, multi-sensor data fusion, multi-target tracking, image segmentation, recursive infer-
ence for force aggregation, inference and learning for multi-scale Markov tree models.
Publications1
Theses
Convex Relaxation Methods for Graphical Models: Lagrangian and Maximum Entropy
Approaches. MIT Doctoral Thesis, 257 pages, August 2008. (3 citations)
http://ssg.mit.edu/group/jasonj
Estimation of GMRFs by Recursive Cavity Modeling. MIT Master’s Thesis, 205 pages,
March 2003. (6 citations) http://ssg.mit.edu/group/jasonj
Journal Articles
JKJ, A. Willsky. A recursive model-reduction method for estimation in Gaussian Markov
random fields. IEEE Transactions on Image Processing, v.17, no.1, pp.70–83, January 2008.
(10 citations) http://ieeexplore.ieee.org
D. Malioutov, JKJ, A. Willsky. Walk-sums and belief propagation in Gaussian graphical
models. Journal of Machine Learning Research, v.7, pp.2031–2064, October 2006. (76
citations) http://jmlr.csail.mit.edu
D. Malioutov, JKJ, M. Choi, A. Willsky. Low-rank variance approximation in GMRF
Models: single and multiscale approaches. IEEE Transactions on Signal Processing, v.56,
no.10, pp.4621–4634, October 2008. (6 citations) http://ieeexplore.ieee.org
1
Google Scholar citations ∼ 256.
3. V. Chandrasekaran, JKJ, A. Willsky. Estimation in Gaussian graphical models using
tractable sub-graphs: a walk-sum analysis. IEEE Transactions on Signal Processing, v.56,
no.5, pp.1916-1930, May 2008. (17 citations)
http://ieeexplore.ieee.org
M. Choi, V. Chandrasekaran, D. Malioutov, JKJ, A. Willsky. Multiscale stochastic model-
ing for tractable inference and data assimilation. Computer Methods in Applied Mechanics
and Engineering, v.197, pp.3492–3515, August 2008. (9 citations)
http://sciencedirect.com
Conference Papers
S. Kudekar, JKJ, M. Chertkov. Linear programming based detectors for two-dimensional
intersymbol interference channels. Submitted to International Symposium on Information
Theory. (under review) http://arxiv.org/abs/1102.5386
JKJ, P. Netrapalli, M. Chertkov. Learning planar Ising models. (under review)
http://arxiv.org/abs/1011.3494
JKJ, M. Chertkov. A Majorization-Minimization Approach to Design of Power Trans-
mission Networks. 49th IEEE Conference on Decision and Control, December 2010. (2
citations) http://arxiv.org/abs/1004.2285
JKJ, V. Chernyak, M. Chertkov. Orbit-Product Representation and Correction of Gaussian
Belief Propagation. International Conference on Machine Learning, June 2009. (4 citations)
http://arxiv.org/abs/090.3769
http://videolectures.net/jason k johnson
JKJ, D. Bickson, D. Dolev. Fixing convergence of Gaussian belief propagation. Interna-
tional Symposium of Information Theory, July 2009. (10 citations)
http://arxiv.org/abs/0901.4192
JKJ, V. Chandrasekaran, A. Willsky. Learning Markov structure by maximum entropy
relaxation. 11th Inter. Conf. on Artificial Intelligence and Statistics (AISTATS), March
2007. (15 citations) http://www.stat.umn.edu/∼aistat/proceedings/start.htm
JKJ, D. Malioutov, A. Willsky. Lagrangian relaxation for MAP estimation in graphical
models. 45th Allerton Conf. on Communication, Control and Computing, September 2007.
(24 citations)
http://www.csl.uiuc.edu/allerton/archives/allerton07/papers/0250.pdf
JKJ, D. Malioutov, A. Willsky. Walk-sum interpretation and analysis of Gaussian belief
propagation. Advances in Neural Information Processing Systems (NIPS), v.18, pp.579–
586, December 2005. (33 citations)
Selected for ”spot-light” at the conference.
http://books.nips.cc/nips18.html
4. V. Chandrasekaran, JKJ, A. Willsky, Adaptive embedded subgraph algorithms using walk-
sum analysis. Advances in Neural Information Processing Systems (NIPS), v.20, December
2007. (4 citations) http://books.nips.cc/nips20.html
V. Chandrasekaran, JKJ, A. Willsky. Maximum entropy relaxation for graphical model
selection given inconsistent statistics. IEEE 14th Workshop on Statistical Signal Processing
(SSP), pp.625–629, August 2007. (3 citations) http://ieeexplore.ieee.org
D. Malioutov, JKJ, A. Willsky. GMRF variance approximation using spliced wavelet bases.
Inter. Conf. on Acoustics, Speech and Signal Processing (ICASSP), v.3, pp.1101-1104, April
2007. (8 citations) http://ieeexplore.ieee.org
D. Malioutov, JKJ, A. Willsky. Low-rank variance estimation in large-scale GMRF models.
Inter. Conf. on Acoustics, Speech and Signal Processing (ICASSP), v.3, pp.676–679, May
2006. (10 citations)
Student Paper Award.
http://ieeexplore.ieee.org
JKJ, R. Chaney. Recursive composition inference for force aggregation. Proc. of the 2nd
Inter. Conf. on Information Fusion, v.2, pp.1187–1195, July 1999. (15 citations)
Alphatech Joseph G. Wohl Memorial Achievement Award.
http://handle.dtic.mil/100.2/ADA366940
W. Irving, JKJ. SAR-FLIR sensor fusion for ATR with 3D model-based reasoning. Proc. of
the 1998 IRIS National Symposium on Sensor and Data Fusion, March 1998. (1 citation)
Invited Talk
Message-Passing Algorithms for GMRFs and Non-Linear Optimization. Neural Information
Processing Systems, Workshop on Approximate Inference in Continuous/Hybrid Models.
Whistler B.C., Canada, December 2007.
http://intranet.cs.man.ac.uk/ai/nips07
Workshops
A Majorization-Minimization Approach to Design of Power Transmission Networks. Mini-
Workshop on Optimization and Control Theory, Los Alamos, NM, August 2010.
http://cnls.lanl.gov/∼chertkov/SmarterGrids/w sh 10/Talks/Johnson.pdf
Orbit-Product Analysis of (Generalized) Gaussian Belief Propagation. Physics of Algo-
rithms Workshop, Santa-Fe NM, September 2009.
http://cnls.lanl.gov/∼jasonj/poa/abstracts.html#johnson
Proposal
M. Chertkov et al. Optimization and Control Theory for Smart Grids. This proposal was
awarded internal R&D funding for a three-year project at LANL.
http://cnls.lanl.gov/∼chertkov/SmarterGrids/