Home

About a Nick
I am not a  PhD student anymore at Georgia Institute of Technology  (ECE), because I graduated  May 1st 2009. My advisors were David Anderson, Alexander Gray. I have been a member of FastLab . My first venture after graduation was  Analytics 1305. Now I am working on a new venture Ismion Inc. Molham Aref has been a great mentor in the jungle of startup business. In the past I have worked with Ron Schafer and Manos Tentzeris. I am a former member of FastLab. It has been a pleasure spending some time in the ISYE corridors fishing interesting ideas and knowledge from the Gurus of Optimization Nemirovski, Shapiro, Monteiro. A nice collateral of my PhD was a sub-PhD in business and innovation I got from my friend Stelios Kavadias after 8 years of drinking coffee at Starbucks. I finished my undergraduate studies in NTUA (Greece). It was really a pleasure  and an honor for me to work with  Petros Maragos on Fractals and Iterated Function Systems. I also made a very good friend, an exceptional mathematician and a beautiful mind Alex Bacopoulos. I am very worried about what is happening in Greece right now. The only way to make it is to trust the party Stephanos Manos founded called DRASSI.
I spend a lot of time(and money) reading and buying books.  My favorite reviewer from Amazon is  DataGuru.

Personal info:


Publications
  1. Nikolaos Vasiloglou, Emir Pasalic, Loren Williams, "Systems and methods for identifying sets of similar products" Patent US20120265736 A1
  2. Manos Antonakakis, Roberto Perdisci, Yacin Nadji, Nikolaos Vasiloglou,Saeed Abu-Nimeh, Wenke Lee,  David Dagon, "From Throw-Away Traffic to Bots: Detecting the Rise of DGA-Based Malware", to appear in the 21th USENIX Security Symposium, Bellevue, WA, August 8–10, 2012.
  3. Arkadas Ozakin, Alexander Gray, Nikolaos Vasiloglou, "Density Preservartion Maps" to appear in "Manifold Learning Theory and Applications", CRC Press 2011.
  4. Manos Antonakakis, Roberto Perdisci, Wenke Lee, Nikolaos Vasiloglou, David Dagon. "Detecting Malware Domains at the Upper DNS Hierarchy". USENIX Security Symposium, 2011
  5. R. Curtin, N. Vasiloglou, D.V. Anderson, "Learning distances to improve phoneme classification", in Proceedings of the 2011 IEEE International Workshop on Machine Learning in Signal Processing (MLSP 2011), Beijing, China, 2011.
  6. "Isometry and Convexity in Dimensionality Reduction", Nikolaos Vasiloglou, PhD dissertation, Georgia Institute of Technology.
  7. "Hyperkernel Based Density Estimation", Ravi S. Ganti, Nikolaos Vasiloglou and Alexander Gray,  NIPS  2009 workshop on kernel learning
  8. "Learning the  Intrinsic Dimensions of the Timit  Speech Database with Maximum Variance Unfolding", N. Vasiloglou, D V. Anderson,  Alexander G. Gray. to appear in the 13th DSP workshop. 
  9. "Learning Isometric Separation Maps", N. Vasiloglou, A. Gray, D. Anderson. NIPS workshop on kernel learning, available at arxiv.org 
  10. "Non-Negative Matrix Factorization, Convexity and Isometry",  N. Vasiloglou, A. Gray, D. Anderson,  to appear in SIAM data mining 2009, available at arxiv.org 
  11. "Scalable Semidefnite Manifold Learning", N. Vasiloglou, A. Gray, D. Anderson, The 2008 IEEE Machine Learning in Signal Processing, Cancun,  Mexico
  12. "Parameter Estimation for Manifold Learning, Through Density Estimation"
    N. Vasiloglou, A. Gray, D. Anderson, The 2006 IEEE Machine Learning in Signal Processing, Maynooth Ireland
  13. ”Towards high quality region-of interest medical video over wireless networks using motion compensated temporal filtering”
    S. Rao, N. Vasiloglou, The 5th IEEE International Symposium on Signal Processing and Information Technology December 2005, Athens, Greece
  14. Isolated word, speaker dependent recognition under the presence of noise, based on an audio retrieval algorithm” N. Vasiloglou, R.W. Schafer, M.C. Hans, Signals, Systems and Computers, 2004. Conference Record of the Thirty-Eighth Asilomar Conference.
  15. Design and optimization of 3D RF modules, microsystems and packages using electromagnetic, statistical and genetic
    tools [mm-wave interdigitated passband filter application]” N. Bushyager, D. Staiculescu, L. Martin, J.-H. Lee, N. Vasiloglou, M.M Tentzeris. Electronic Components and Technology, 2004. ECTC ’04. Proceedings
  16. Fast hybrid electromagnetic/statistical approach for design and optimization of RF systems and packages” N. Bushyager, D. Staiculescu, L. Martin, J.-H. Lee, N. Vasiloglou, M.M Tentzeris. M.M. Advanced Packaging Materials: Processes, Properties and Interfaces, 2004.
  17. Investigation of the Effect of Fractal Shapes on the Broadband Behavior of One-Dimensional Optimized Antennas”,N.
    Vasiloglou, D.Staiculescu and M.M.Tentzeris, Proc. of the 2004 URSI Symposium, p.69, Monterey, CA, June 2004.
  18. Lossless audio coding with MPEG-4 structured audio” N. Vasiloglou, R.W. Schafer, M.C. Hans, Web Delivering of Music, 2002 Proceedings.
  19. Spectrum of Fractal Interpolation Functions, N. Vasiloglou, P. Maragos, summary of my undergrad thesis, still needs some polishing to get published.






Indust
rial Interests
  • Large scale distributed machine learning.
  • My dream is to apply machine learning in the music recommendation/radio industry. I do have ideas, I am prototyping them on the side. I am looking for partners
  • Startups are always attractive to me

Work Experience
  • HP-Labs, summer 2002
  • Google, summer 2004, where I met Shahid Choudhry a very interesting engineer that inspired me a lot
  • Google, summer 2006, I did improve my coding  skills a lot
  • Admob 2008, where I met Ayman Farahat a very sharp researcher
  • I founded with Alex Gray Analytics 1305. It is a startup for large scale machine learning applications
  • I just joined Ismion Inc a machine learning company focusing on simple solutions.
  • Along with Manos Antonakakis we have done some very interesting work on modeling Bots specially the ones that use DGAs (Domain Generating Alogrithms) at Damballa. Machine learning seems to be winning the game there
  • My long term collaboration with LogicBlox and Predictix starts generating fruits on some large scale  retail problems.
  • I feel lucky I got the opportunity to work with Lee Edlefsen at Revolution Analytics. As far as I have seen so far it is the most powerful numerical engine for big data. Lee has over 30 years experience in numerical software engineering. 
  • I have also been working with LexisNexis HPCC platform. My experience so far indicates that it is the best tool for big data. I am an official partner with them and so far our collaboration seems to go very well.
  • I found a great team to work with at Tapad.

Academic Interests
  • I have been looking lately into alternative factorizations based on minmax algebra.
  • Approximation of similarities with metrics. In general I think the cornerstone of machine learning is fast computation of similarities.
  • A very exciting but forgotten are of machine learning that I belive has great potential is
     polynomial neural networks.
  • As a former member of FastLab I am deeply into kd-trees ball-trees and everything that comes from Andrew Moore's heritage
  • I participated actively in the FastLib C++ development project for Scalable Machine Learning.
  • Large Scale non-linear optimization is another hot area that I have worked on. I am particularly interested in Semidefinite programming (Convex and Non convex). It has great potential in machine learning and a lot of algorithms have been presented on small scale.
  • Global Optimization (GoP) is the hot potato for me now. I like the branch and bound techniques, they have been quite unsuccessful in the nearest neighbor problems and it seems to me that nobody has tried GoP in machine learning problems yet.
  • The world of numerical computing would have been much better if computers were supporting interval arithmetic. I contest that this can be the next revolution in computing after GPUs.
  • Generalized Geometric Programming along with GoP can solve many interesting machine learning problems
  • My biggest love is Chaos and Fractals and this is because of Michael Barnsley and his Fractals everywhere. Plenty of ideas are rambling into my mind. After my PhD is over I will revisit some of the great ideas in Barnsley's book. I would also like to take a closer look at his more recent book SuperFractals.
Subpages (1): What is keeping me busy
Ċ
main.pdf
(10939k)
Nikolaos Vasiloglou II,
Jan 14, 2012, 1:05 PM
Ċ
Nikolaos Vasiloglou II,
Jan 14, 2012, 1:01 PM
Comments