Fangjian Guo

(a.k.a. Richard Guo / 郭方健 / richardkwo)
My photo 

Department of Computer Science,
Duke Univeristy,
Durham, NC 27708, USA.


This site is mirrored at

Contact

Duke CS logo Email

  • Academic:

  • Personal:

GitHub    LinkedIn    Twitter    Physical Address

Research Interests

  • Scalable machine learning: parallel MCMC, stochastic/subsampling-based algorithms, relaxations

  • Bayesian structural modeling: sparsity, low rank, networks, etc.

  • Nonparametrics Bayes

  • Stochastic processes: point process modeling for social interactions & human behaviors

  • Connecting learning and statistical mechanics: Monte Carlo, belief propagation, phase transition, etc.

Short Bio

I am a graduate student in the Department Computer Science, Duke University. I am working on machine learning methodology with Prof. Katherine Heller in the Duke Statistical Machine Learning Group. I also work with Prof. David Dunson.

Before that, I got my bachelors degree in computer science from University of Electronic Science and Technology of China (UESTC) in Chengdu, Sichuan. (Chengdu is my 2nd hometown where I lived for 7 years.) As part of my undergraduate studies, I worked with Prof. Tao Zhou on a few research topics related to human behaviors and social networks, including ranking algorithms, modeling social network evolution and statistical properties of human behaviors.

Even before that, I grew up in Deyang, Sichuan, in the community of a huge state-owned factory (heavy industrial parts still evoke my childhood belongingness), where most of my family members are/were working. Then I attended the century-old Chengdu No. 7 High School.

While being a CS major, physics has also been my long-time fascination. I am interested in research that explores the deep connections between machine learning and statistical physics.

What's new

  • [2015/06]   Summer intern at Microsoft Research Asia with David Wipf working on Bayesian sparse and low-rank estimation algorithms.

  • [2015/06]   New paper on parallelizing MCMC with random trees.

  • [2015/04]   Developing a much faster version of “Bayesian Echo Chamber” that is scalable to large corpus of conversation transcripts, movie subtitles, screenplays, dramas, etc.

  • [2015/02]   Working on a novel inference algorithm for DP mixtures that is fast and takes care of uncertainty.

  • [2014/12]   Looking for a research intern over the upcoming summer. Interested in R&D positions in machine learning, NLP and analytics. This is a brief summary of my ongoing research.

  • [2014/11]   Working with Samuel Wang on novel ensemble learning algorithms powered by DPP.

  • [2014/06]   Will be hanging around ICML 2014 at Beijing. Shoot me an email if you also happen to be there and interested in chatting with me (I'd be happy to meet u!).

  • [2014/06]   Submitted something to NIPS.

  • [2014/03]   Working on modeling the dynamics of interpersonal communication with Hawkes process and other Bayes nonparametric methods. Meanwhile, started to learn playing guitar.

  • [2014/03]   Declared my advisor! Glad to be (one of) her first student. :)

  • [2013/08]   I joined Duke as a PhD student in computer science! A fresh blue devil now!

  • [2013/08]   The paper below appeared on Physica A.

  • [2013/07]   I attended STATPHYS 25 in Seoul, Korea. The place and people are lovely!

  • [2013/06]   I defended my bachelor's thesis and graduated from UESTC with the highest distinction. Farewell, my dear fella and good old days!

  • [2013/03]   I made an introductory lecture on mathematical modeling for 1st-year and 2nd-year undergraduates in my home university. Slides can be found here.

  • [2013/03]   I am back from my visiting at City University of Hong Kong. I am continuing the new studies on the “viral marketing” in social networks growth in collaboration with Prof. Jonathan Zhu.

  • [2013/01]   New updates on my project investigating the relation between memory (autocorrelation) and power-law.

  • [2012/12]   I feel honored to be awarded as one of the Outstanding Students in my university, the highest honor for UESTC graduates. (10 out of 4,500 undergraduates in 09’)

  • [2012/11]   I am currently working on the project of the memory of power-law random series, see my projects for more information.

Hobbies

  • Typography & typesetting

  • Graphical design

  • Classical poetry (mostly Chinese)

  • Badminton

  • Reading The Hitchhiker's Guide to the Galaxy and other stuff of this flavor!

  • Physics (and yes, it is a hobby): I find this very engaging.

  • Exploring other interesting stuff (and yes, it is a recursion)