Sung-eok Jeon, Ph.D.

Short Bio.:
Sung-eok Jeon received B.S. degree from EE Department at Yonsei University, Seoul, Korea; M.S. degree from EE Department, KAIST, Daejeon, Korea; and Ph.D. degree from the School of ECE at Georgia Institute of Technology, Atlanta GA. He worked at Microsoft, Seattle, WA, from Jan/2008 to Aug/2015 about Windows system reliability analysis
(large system reliability analysis and probabilistic modeling; applying machine learning models on data analysis and modeling), data mining, and software engineering on data pipelines, user-mode debugging, and bug-filing systems.
He worked at Facebook from Aug/2015 to Oct/2017 (filed over 10 patents with Facebook patent lawyers and co-workers). He is now at Uber since Oct/2017. 
His research interests are statistical modeling of large and complex systems based on the probabilistic graphical models in machine learning;deep-learning on complex data with variant neural networks; message-passing algorithms over graphical models; data modeling/analysis/mining and information retrieval with machine learning approaches. 
 
Career Experience:
  • Uber, Oct/2017~Now
  • Facebook, Aug/2015~Oct/2017
  • Software Development and Data Analysis, Microsoft Inc., Redmond, WA, Jan/2008-Aug/2015
    Various analysis, modeling and data mining about Windows system/features/processes reliability and behaviors with machine learning approaches.
    Deterministic/probabilistic modeling, feature engineering, predictive modeling, causal analysis, and etc.
    Parameter learning with message passing and Probabilistic Graphical models (MRF, HMM, etc).
    Data mining and statistical analysis on the reliability of Windows Systems, Processes, and Services. 
    User-mode Debugging
  • Graduate Research and Teaching Assistant, Georgia Institute of Technology, Atlanta, GA
    Researches on the near-optimality on the randomized and distributed self-configuration of wired and wireless networks.
  • Graduate Research and Teaching Assistant, KAIST, Taejeon, Korea

Awards and Recognition:
  • 5 Times Academic Excellence Awards from Yonsei University President.
  • Upper 4.0/4.0, Early Graduation in 3.5 years, Yonsei University. 
  • Listed in Marquis Who's Who in America since 2007 (Computer Scientist).
  • Filed over 10 patents at Facebook
 
Skills:
  • C#/C++/C, Python, SQL, R, Matlab, Java
  • Data modeling (predictive modeling, and statistical modeling), analysis, and mining with machine learning approaches (with Python, R, SQL) 
  • Learning based on probabilistic Graphical Models (MRF, MLP, RBM), and message-passing over probabilistic graphical models
  • Linux/Unix/Windows
  • Debugging (mostly about user-mode)
  • Computer Network Theories (Queuing, Routing, Scheduling)/Protocols/Test-beds/Discrete Event Driven Network Traffic Simulations/Resource Optimization
  • Statistical/Deterministic Optimization

Selected Publications:  
  1. S. Jeon, and et.al, Multiple technical reports about Windows System Reliability and Usage Analysis, Jan/2008~now.
  2. S. Jeon, and C. Ji, "Distributed Configuration Management of Wireless Networks: Markov Random Field and Near-Optimality," IEEE Trans. Signal Processing, vol. 58(9), pp. 4859-4870, Sep., 2010.
  3. P. Li, R. Kivett, Z. Zhan, S. Jeon, N. Nagappan, B. Murphy, and A. Ko, "Characterizing the Differences Between Pre- and Post- release Versions of Software," in Proc. of IEEE International Conference on Software Engineering (ICSE), April, 2011.
  4. S. Jeon, and C. Ji, "Nearly-Optimal Distributed Configuration Management Using Probabilistic Graphical Models," in Proc. of IEEE MASS, RPMSN Workshop, 2005.
  5. S. Jeon, and C. Ji, "Graphical Models for Self-Configuration of Ad-hoc Wireless Networks," in Proc. of Snowbird Learning Workshop, 2005. 
  6. S. Jeon, and C. Ji, "Role of Machine Leaning in Self-Configuration of Ad-hoc Wireless Networks," in Proc. of ACM SIGCOMM, MineNet Workshop, 2005.
  7. S. Jeon, "Topology Aggregation Method for Multiple Link Parameters," Journal of Computer Communications, Elsevier, Feb. 2006.
  8. S. Jeon, "Redundant Protection Problem in the Hierarchical GMPLS Networks," in Proc. of IEEE LCN, 2005.
  9. S. Jeon, and R.T. Abler, "Formulation and Optimization of the Connection Preemption," Journal of Computer Communications, Elsevier, 2005.
Subpages (1): Finding Ego: