Danny Bickson's Homepage

I am now with GraphLab Inc.
4000 Mason Rd, Seattle WA
View Danny Bickson's profile on LinkedIn

News:


  • May 2013: GraphLab raises 6.75M$ to form a company around its open source project.

    • Reseach Interests

    My research combines large scale distributed algorithms design and their deployment, spanning both the theoretical and applied aspects of large scale computing. I have strong theoretical background in distributed computing, message-passing algorithms, linear algebra, machine learning, numerical iterative methods, probabilistic graphical models, convex optimization and information theory. From an applied perspective, I have extensive experience with the design and building large-scale systems for solving practical problems on massive datasets using different computing platforms: Cloud computing platforms like Amazon EC2, IBM Bluegene supercomputer, PlanetLab testbed, Sensor networks, GPUs and other parallel/distributed environments. My most recent work, GraphLab, involves the design and implementation of a parallel machine learning framework, targeted for iterative and potentially asynchronous algorithms.

    Education

    Postdoctoral Researcher Machine Learning Dept, Carnegie Mellon University (August 2009-Aug 2013) Member of the SELECT (sense, learn, act) lab. Sponsors: Prof. Carlos Guestrin (CMU) and Prof. Joseph Hellerstein (UC Berkeley). Working on building GraphLab - a distributed machine learning framework for supporting asynchronous computation on sparse graphs.

    Postdoctoral Researcher IBM Haifa Research Lab (March 2008-July 2009) Member of the messaging group. Worked on applying machine learning algorithms to practical network problems.

    Ph. D. Computer Science, Hebrew University of Jerusalem. 2003--2008
    Advisors: Prof. Danny Dolev and Prof. Dahlia Malkhi.
    Main research interest: distributed algorithms, Peer-to-Peer and social networks, machine learning, message-passing algorithms, numerical iterative methods, information theory, graphical models. Thesis title: Gaussian belief propagation: Theory and application. arxiv

    M. Sc. Computer Science, Hebrew University of Jerusalem. 2001--2003
    Advisor: Prof. Dahlia Malkhi. Member of the DANSS (Distributes Algorithm, Networking and Secure Systems) group. Thesis title: Steganographic communications in the Gnutella network.

    B. Sc. Computer Science, Hebrew University of Jerusalem. 1996--1998.

    Professional Activities

  • Program committee of GRADES 2014: Graph Data-management Experiences and Systems, in conjunction with ACM SIGMOD/PODS 2014.
  • Program committee of 2014 Workshop on Big Graph Mining (BGM), in conjunction with WWW conference 2014.
  • Organizer of Large Scale Recommender System workshop (LSRS 2013) in conjunction with Recsys conference, 2013.
  • Organizer of the 2nd GraphLab workshop, 2013.
  • Organizer of the 1st GraphLab workshop, 2012.
  • Program committee of Journal of Machine Learning Research, Special Issue on KDD CUP 2011.
  • Program committee of NIPS 2011 Big Learning Workshop, 2011.
  • Program committee of ICML 2011 - 28th International Conference on Machine Learning, 2011.
  • Program committee of the 11th IEEE Peer-to-Peer Computing Conference, 2011.
  • Program committee of the 10th IEEE Peer-to-Peer Computing Conference, 2010.
  • Program committee of the 9th IEEE Peer-to-Peer Computing Conference, 2009.
  • Program committee of the Global Internet Symposium (GI 06'), Apr. 2006, Barcelona.
  • Organizing committee of the 2nd European PlanetLab meetings, Oct.2005, Switzerland.
  • Reviewer for journals: IEEE Tran. on Multimedia, IEEE Tran. on Sensor Networks, IEEE Tran.on Networking, IEEE Tran.on Automatic Control, IEEE Tran. on Signal Processing, IEEE Tran.on Wireless Communications, EURASIP Journal of Wireless Communications and Networking, Journal on Selected Areas in Communications.

    My blog!

    is here

    Grants

  • Evergrow (Ever Growing Global Scale Free Networks) EU project (FP6, ICT), 5.6M Euro, assistant coordinator. Helped coordinate a joint research project of 26 European organizations, including identifying research goals and tracking progress, arranging periodical meetings, annual reporting and coordination.
  • HiPerDNO (High Performance Computing Technologies for Smart Distribution Network Operation) EU FP7 proposal, 4.4M Euro, initiated and wrote IBM HRL's part in this funded proposal.

    Fellowships and Awards

  • WCSD Challenge 2013, winner of 3rd place, 2013.
  • ACM KDD CUP, winner of 4th place in track2 (out of 192 teams), 2012
  • 5th place in ACM KDD CUP contest 2011 (out of more than 1000 groups).
  • Leibniz Fellowship for the academic year 2005.
  • Garusso Prise for the academic year 2004.

    Source Code

  • GraphLab parallel machine learning framework
  • Gaussian belief propagation Matlab Toolbox
  • Linear Characteristic Graphical Models (Linear-Stable) Matlab Toolbox

    Teaching Experience

  • 2004-2008 - The Hebrew University of Jerusalem. Instructor of Digital Communications in the Modern World course 2004-5, 2005-6, 2006-7, 2007-8.
  • 2007 - Interdisciplinary Center (IDC), Herzlyia. Instructor of ACNS (Advanced Computer and Network Security) course 2007.
  • 2001-2005 - Hadassah Academic College, Jerusalem. Instructor of Intro2cs2 course 2004-5, Digital Design course 2002-3, Introduction to Computer Communications course 2001-2.

    Publications

    Journals

  • D. Bickson, D. Baron, Alex T. Ihler, H. Avissar and D. Dolev. Fault identification via non-parametric belief propagation. In IEEE Tran. on Signal Processing, vol. 59, no. 6, pp. 2602-2603, June 2011 arxiv bibtex
  • D. Bickson, T. Reinman, D. Dolev and B. Pinkas. Peer-to-Peer secure multi-party numerical computation facing malicious adversaries. In Peer-to-Peer Networking and Applications Journal, Vol. 3, No. 2, June 2010, Springer, pp. 129-144. arxiv bibtex
  • D. Bickson and D. Malkhi. A unifying family of rating users and data items in Peer-to-Peer and social networks. In Peer-to-Peer Networking and Applications journal, Vol. 1, No. 2, Sept. 2008, Springer, pp. 93-103. pdf bibtex

    Refereed Conferences

  • J. Gonzalez, Y. Low, H. Gu, D. Bickson and C. Gustrin. PowerGraph: Distributed Graph-Parallel Computation on Natural Graphs. In the 10th USENIX Symposium on Operating Systems Design and Implementation (OSDI '12).
  • Y. Low, J. Gonzalez, A. Kyrola, D. Bickson, C. Guestrin, and J.M. Hellerstein. Distributed GraphLab: A Framework for Machine Learning in the Cloud. PVLDB 5(8):716-727, 2012.
  • J. Bradley, A. Kyrola, D. Bickson and C. Guestrin. Parallel Coordinate Descent for L1-Regularized Loss Minimization. In the 28th International Conference on Machine Learning (ICML 2011), to appear. arxiv
  • L. Song, A. Gretton, D. Bickson, Y. Low and C. Guestrin. Kernel Belief Propagation. In the 14th International Conference on Artificial Intelligence and Statistics (AISTATS) 2011. arxiv bibtex
  • D. Bickson and C. Guestrin. Inference with multivariate heavy-tails in linear models. In Neural Information Processing Systems (NIPS) 2010, Vancouver, Canada, Dec. 2010. arxiv
  • Y. Low, J. Gonzalez, A. Kyrola, D. Bickson, C. Guestrin and J. Hellerstein. GraphLab: A New Framework for Parallel Machine Learning. In the 26th Conference on Uncertainty in Artificial Intelligence (UAI), Catalina Island, USA, 2010. arxiv bibtex
  • D. Bickson, H. Avissar, A. Ihler and D. Dolev. A Low density lattice decoder via non-parametric belief propagation. In the 47th Annual Allerton Conference on Communication, Control and Computing, Allerton House, Illinois, Sept. 2009 pp. 439-446. arxiv bibtex
  • J. K. Johnson, D. Bickson and D. Dolev. Fixing convergence of Gaussian belief propagation algorithm. In Proc. of the International symposium on Information Theory (ISIT), Vol. 3, June 28 - July 3, 2009, Coex, Seoul, Korea, pp. 1674-1678. arxiv bibtex
  • D. Bickson, Y. Tock, A. Zymnis, S. Boyd and D. Dolev. Distributed large scale network utility maximization. In Proc. of the International symposium on Information Theory (ISIT), Vol. 2, June 28 - July 3, 2009, Coex, Seoul, Korea, pp. 829-833. arxiv bibtex
  • E. N. Hoch, D. Bickson and D. Dolev. Self-stabilizing numerical iterative computation. In Proc. of the 10th International Symposium on Stabilization, Safety, and Security of Distributed Systems, Volume 5340/2008, Detroit, Nov. 2008, pp. 79-93. arxiv bibtex Journal version arxiv
  • D. Bickson, O. Shental, and D. Dolev. Distributed Kalman filter via Gaussian belief propagation. In proc of the 46th annual Allerton Conference on Communication, Control and Computing, Urbana-Champaign, IL, Sept. 2008, pp. 628-635. arxiv bibtex
  • D. Bickson, O. Shental, Y. Tock and D. Dolev. Polynomial linear programming with Gaussian belief propagation. In proc. of the 46th annual Allerton Conference on Communication, Control and Computing, Urbana-Champaign, IL, Sept. 2008, pp. 895-901. arxiv bibtex
  • D. Bickson, D. Dolev, G. Bezman and B. Pinkas. Peer-to-Peer secure multi-party numerical computation. In the 8th International IEEE Peer-to-Peer Computing 2008, Aachen, Germany, Sept. 2008, 257-266. arxiv bibtex
  • O. Shental, D. Bickson, P. H. Siegel, J. K. Wolf, and D. Dolev. Gaussian belief propagation solver for systems of linear equations. In proc. of the IEEE Int. Symp. on Inform. Theory (ISIT), Toronto, Canada, July 2008, pp. 1863-1867. arxiv bibtex
  • D. Bickson, O. Shental, P. H. Siegel, J. K. Wolf, and D. Dolev. Gaussian belief propagation based multiuser detection. In proc. of the IEEE Int. Symp. on Inform. Theory (ISIT), Toronto, Canada, July 2008, pp. 1878-1882. arxiv bibtex
  • T. Anker, D. Bickson, D. Dolev and B. Hod. Efficient Clustering for Improving Network Performance in Wireless Sensor Networks. In European Wireless Sensor Networks (EWSN 08), Bologna, Italy, Feb. 2008. pdf bibtex
  • E. Jaffe, D. Bickson and S. Kirkpatrick. Everlab - A production platform for research in network experimentation and computation. In the 21st Usenix Large Installation System Administration Conference (LISA '07), Dallas, Texas, Nov. 2007. pdf bibtex
  • D. Bickson, D. Dolev, O. Shental, P. H. Siegel and Jack K. Wolf. Linear Detection via Belief Propagation. In the 45th Annual Allerton Conference on Communication, Control, and Computing, Allerton House, Illinois, Sept. 2007. pdf bibtex
  • D. Bickson, D. Malkhi and L. Zhou. Peer-to-Peer Rating. In proc. of the 7th IEEE Peer-to-Peer Computing Conference, Galway, Ireland, Sept. 2007, pp. 211-218. pdf bibtex
  • D. Bickson and D. Malkhi. The Julia Content Distribution Network. In the 2nd Usenix Workshop on Real, Large Distributed Systems (WORLDS '05), Dec. 05', SF, CA. pdf bibtex
  • I. Abraham, A. Badola, D. Bickson, D. Malkhi, S. Maloo and S. Ron. Practical Locality-Awareness for Large Scale Information Sharing. The 4th Annual International Workshop on Peer-To-Peer Systems (IPTPS '05), Itacha, NY, Feb. 2005. pdf bibtex
  • D. Bickson, D. Malkhi and D. Rabinowitz. Efficient Large Scale Content Distribution. In proceedings of the 6th Workshop on Distributed Data and Structures (WDAS'2004), Lausanne, Switzerland, July 2004. ps bibtex

    Other publications

  • Qiang Yan, XingXing Wang, Qiang Xu, Dongying Kong, Danny Bickson, Quan Yuan, and Qing Yang (2013). "Predicting Search Engine Switching in WSCD 2013 Challenge." Workshop on Web Search Click Data (WSCD) 2013. pdf
  • Xingxing Wang, Shijie Lin, Dongying Kong, Liheng Xu, Qiang Yan, Siwei Lai, Liang Wu, Guibo Zhu, Heng Gao, Yang Wu, Danny Bickson, Yuanfeng Du, Neng Gong, Chengchun Shu, Shuang Wang, Fei Tan, Jun Zhao, Yuanchun Zhou, Kang Liu. Click-Through Prediction for Sponsored Search Advertising with Hybrid Models. In ACM KDD CUP workshop 2012.
  • Yao Wu, Qiang Yan, Danny Bickson, Yucheng Low, Qing Yang. Efficient Multicore Collaborative Filtering. In ACM KDD CUP workshop 2011. arxiv.
  • Y. Low, G. Gonzalez, D. Bickson, A. Kyrola and C. Guestrin. Machine learning in the cloud using GraphLab. In learning on cores, clusters and clouds (LCCC) workshop, NIPS 2010, Dec. 2010, Vancouver, Canada. Full tech report: arxiv.
  • D. Bickson, H. Hoch, N. Nir and Y. Tock. A hybrid unicast-multicast infrastructure for efficient publish subscribe in enterprise networks. In proc. of the 3rd Annual Haifa Experimental Systems Conference (SYSTOR 2010), pp. 1-7, Haifa, Israel, ACM. arxiv bibtex
  • O. Shental, D. Bickson, P. H. Siegel, J. K. Wolf, and D. Dolev. A message-passing solver for linear systems, in Proc. Information Theory and Applications (ITA) Workshop, San Diego, CA, USA, January 2008, p. 360. pdf bibtex
  • D. Bickson, D. Dolev and E. Yom-Tov. Gaussian belief propagation solver for large scale support vector machines. In the 5th European complex systems conference (ECCS '08), Jerusalem, Sept. 2008. arxiv bibtex
  • D. Bickson, G. Gershinsky, E. Hoch and K. Shagin. A statistical approach to monitoring of soft-real time distributed systems. Nov. 2008. Tech Report. arxiv bibtex
  • D. Bickson, D. Dolev and E. Yom-Tov. Solving Large Scale Kernel Ridge Regression using a Gaussian Belief Propagation Solver (poster). In NIPS Workshop on Efficient Machine Learning, 2007. pdf bibtex
  • D. Bickson and R. Borer. BitCode: A BitTorrent Clone using Network Coding. In the 7th IEEE Peer-to-Peer Computing. Galway, Ireland, Sept. 2007, pp. 231-232. pdf bibtex
  • D. Bickson, D. Dolev and E. Yom-Tov. Solving Large Scale SVM using Gaussian Belief Propagation Leibniz Center TR-2007-117, School of Computer Science and Engineering, The Hebrew University, 2007.
  • K. Aberer, D. Bickson, D. Dolev, M. Hauswirth and Y. Weiss. Indexing data-oriented overlay networks using belief propagation. In the 7th Workshop of distributed algorithms and data structures (WDAS 06'), Jan. 06', SF, CA. pdf bibtex
  • K. Aberer, D. Bickson, D. Dolev, M. Hauswirth and Y. Weiss. Extended Data Indexing in Overlay Networks using the Belief Propagation Algorithm. In the Peer-to-peer Data Management in the Complex Systems Perspective Workshop (ECCS 05'). pdf (Invited paper). Nov. 2005, Paris.
  • D. Bickson, D. Dolev and Y. Weiss. Modified Belief Propagation for Energy Saving in Wireless and Sensor Networks. Leibniz Center TR-2005-85, School of Computer Science and Engineering, The Hebrew University, 2005. pdf
  • Y. Kulbak and D. Bickson. The eMule Protocol Specification. Leibniz Center TR-2005-3, School of Computer Science and Engineering, The Hebrew University, 2005. pdf bibtex
  • D. Bickson and D. Malkhi. Locality-Aware Content Distribution. Leibniz Center TR-2004-52, School of Computer Science and Engineering, The Hebrew University, 2004. pdf
  • D. Bickson and D. Malkhi. A Survey of Privacy in Peer-to-Peer Networks. Leibniz technical report TR-2003-67, The Hebrew University of Jerusalem. ps

    Patents

  • US Patent 8,139,656: Method and system for linear processing of an input using Gaussian Belief Propagation, with O. Shental, P. H Siegal and D. Dolev. March 2010 (granted)
  • US Patent 8,621,016 : Adaptive di¿erential propagation of SOAP messages. With G. Gershinsky and K. Shagin. (granted)
  • US Patent 20,110,078,291: Distributed performance monitoring in soft real-time distributed systems. With G. Gershinsky and K. Shagin, 2011. (application)

    Some of my travels!


    View Travel Map in a larger map
    Algorithmique distribu'e, du 4 mai au 8 mai 2003
    Porquerolles, France.     		WDAS 2004 conference, 8-9 July 2004, EPFL,
    Lausanne, Switzerland.     		Evergrow SP 3 meeting, Louvain – la –Neuve, Belgium, June 2004
    Photo taken in Nice by Yaron Weinsberg. Photo taken in EPFL Lausanne by Prof.
    Manolis Koubarakis     		Photo taken in Lueven by Prof. Danny Dolev.
    Evergrow SP3 meeting, Kista, Sweden, August 2004 Evergrow SP3 content distribution meeting, EPFL,
    Switzerland, March 2005     		Networks and Algorithms Workshop, Helsinki,
    Finland, June 2005
    Photo taken by Sameh El-Ansary (SICS). Photo taken in Helisnki by Prof. Danny Dolev. Photo taken by Dr. Stephan Boukong (Aston).
    2nd European PlanetLab meeting, EPFL, Switzerland,
    October 2005     		Evergrow SP3 meeting, EPFL, Switzerland, July 2005 Evergrow 2nd Workshop, Sastaholm, Sweden, Dec. 2005
    Photo taken in Luzern by Assaf Ben David Photo taken in Laussane by Prof. Danny Dolev. Photo taken by Elliot Jaffe in Sastaholm.
    P2P Data Management in the Complex
    Systems Perspective ECCS 05' Workshop,
    Paris, Nov. 2005     		WDAS 2006, CA, Jan 2006 Evergrow EC Review, University of Rome,
    Italy, March 2006
    Photo taken by my lovely wife, Ravit Photo taken by Yaniv Hen, NY Photo taken by prof. Thomas Schwartz in Santa Clara, CA
    Evergrow EC Review, University of Rome,
    Italy, March 2006     		Evergrow SP3+4 meeting, ISI, Torino,
    Jan 2006     		Intern In Microsoft Research, Mountain
    View, CA, July 2006
    Photo taken by Elliot Jaffe, Italy Photo taken by Coby Fernandess in Torino, Italy Photo taken by my wife Ravit in Yosemite, CA
    Intern In Microsoft Research, Mountain View,
    CA, June 2006     		Evergrow meetings, Rome, December 2006. Intern In Microsoft Research, Mountain View,
    CA, Aug 2006
    Photo taken by my wife Ravit in Monterey, CA Photo taken by Elliot Jaffe Photo taken by my father in San Jose, CA
    Evergrow meeting, Stockholm, March 2007. Evergrow EU review, Zurich, March 2007. Peer-to-Peer computing 2007, Galway, Ireland,
    Sept. 2007.
    Photo taken by prof. Scot Kirkpatrick. Photo taken by Udi Weinsberg Photo taken by Boris Meijas, UCL.
    Evergrow meeting, Athens, June 2007. Huminet meeting, Frankfurt, February 2008. Evergrow EU meeting, Torino, Dec. 2007.
    Photo taken by Udi Weinsberg Photo taken by Lev Muchnik Photo taken by Udi Weinsberg
    NORDITA physdis workshop, Stockholm, May 2008. Evergrow EU final review, Torino, March 2008. ISIT 2008 conference, Toronto, Canada, July 2008.
    Photo taken by Udi Weinsberg Photo taken by Udi Weinsberg Photo taken by John Doe
    Vacation in Antalya (Hadrianus Gate), Aug. 2008. P2P conference, Aachen, Germany, Sept. 2008. Visit to Los Alamos National Lab, Los Alamos
    New Mexico, Sept. 2008
    Photo taken by my 5 years old (!) son Itay. Photo taken by Boriss Meijas. Photo taken by Dr. Jason K. Johnson.
    46th Allerton conference, Allerton House, Montecillo,
    Illinois, Sept. 2008.     		CMU visit, Pittsburgh, Feb. 2009. Berkeley Visit, Berkeley, CA, Feb. 2009.
    Photo taken by John Doe. Photo taken by Jhon Doe. Photo taken by Jhon Doe.
    IEEE Int. Symp. on Info. Theory. Seoul,
    July 2009.     		47th Allerton conference, Allerton House,
    Montecillo Illinois, Sept. 2009.     		Simmon's farm, PA, Oct. 2009.
    Photo taken by John Doe. Photo taken by Dror Baron. Photo taken by my son, Dor.
    Visiting Cleveland, Ohio, May 2010. Me and my canoe, PA, June 2010. Visiting New-York, July 2010.
    Photo taken by my son, Dor. Photo taken by my son, Dor. Photo taken by my son, Dor.
    UAI 2010, July 2010, Catalina Island, CA. Vacation in the Dead-Sea, July 2010, Israel. Vacation in Jerusalem, Mahne Yehuda Market, Aug. 2010,Israel
    MURI project meetings, Aug. 2010, Salt Lake City, Utah. Vacation in Washington D.C., Nov. 2010. NIPS 10 conference, Dec. 2010, Vancouver, CA.
    Photo taken by Prof. Rohit Negi, CMU. Photo taken by my son, Dor. Photo taken by Aapo Kyrola, CMU.
    Visit to Israel, Dec. 2010. Snow in Pittburgh, Jan 2010. Niagra Falls, April 2011.
    Photo taken by my father Photo taken by my son, Dor. Photo taken by Hadas P.
    Visit to the Bay Area, GeekSessions, May 2011. NY Visit, May 2011. Ohio State Univ Visit, Nov 2011.
    Photo taken by Shon Burton Photo taken by Ido Rekem Photo taken by Josh Ash
    NIPS 2011 Workshop - Sierra Nevada, Spain UCL Visit, London, Feb 2012 Pittsburgh visit, Feb 2012
    Photo taken By John Doe Photo taken by Arthur Gretton Photo taken by Doron Naveh
    Lawa Meetings, Budapest, March 2012 Visiting CMU, April 2012 Gamla Nature Reserve, Israel April 2012
    Photo taken by Scott Kirkpatrick Photo taken by Dr. Naveh Doron Photo taken by my son Dor
    GraphLab Workshop, San Fransisco, July 2012 ACM KDD, Beijing, Aug 2012
    Photo taken by Guy Itzhaki Photo taken by Justin Yan
    Free PageRank Checker