My Profile

Aleksandr Drozd

wherever you go -
there you are

Home | Projects | PGP Key

Profile

Aleksandr Drozd (Ph.D.) is a post-doctoral researcher at the Global Scientific Information and Computing Center, Tokyo Institute of Technology. His research interests lie at the intersection of High Performance Computing and intelligent data processing: data analytics, machine learning and natural language processing.


Education

PhD course

2010-2014

Tokyo Institute of Technology, Graduate School of Information Science and Technology, Tokyo. Thesis title: "Memory-Conscious Optimizations for Sorting and Sequence Alignment for Massively Parallel Heterogeneous Architectures".

Master degree

2000-2005

Moscow State University. Thesis title: "Semantic Pseudo-Code: Approach to Meaning-Base Search."

Academic Profile

My main interest is on the intersection of the high performance computing (HPC) and intelligent data processing in various applied tasks. The areas in which I have worked (both by myself and in collaboration with experts in these areas) include:
  • Artificial intelligence (artificial life modeling, swarming behaviour, social simulations)
  • Natural language processing (high-performance construction of word embedding and their use in information retrieval)
  • Computational biology (high-performance processing of big genomic data on accelerators)

Relevant Skills

I am a passionate programmer who does a fair amount of coding for research and sometimes for fun.
  • Coding/Software Development: My experience as a developer of commercial software gave me such skills as object oriented design, patterns and development processes.
  • C, C++ (including C++14 standard), along with such libraries and tools for parallel programming as CUDA, OpenMP, MPI, OpenCL, TBB, etc for performance-critical parts.
  • Python for everything else: high level scripting, quick prototyping and such. Being open-source enthusiast I’m trying to contribute back to the Python ecosystem by submitting code to the Python core libraries such as Pandas.
  • I have experience with databases (SQL and noSQL), web technologies and version control systems, computer algebra and publishing systems.
  • I also use machine machine learning extensively - from basic statistical analysis methods to artificial neural networks.
  • UNIX system administration.

Languages

  • Russian - Native
  • English - Fluent
  • Japanese - Intermediate
  • Ukrainian - Intermediate
  • French - Basic

Work Experience

Postdoctoral researcher

2014-onwards, Global Scientific Information and Computing Center(GSIC), Tokyo Institute of Technology

Teaching assistant

2005-2010, Moscow State University (Sevastopol Branch)

Teaching courses on parallel data processing, computer graphics and system programming.

Software developer

2006-2009, Outsourcing Ukraine

C++/C# programming, software design.

Software developer

2005-2006, Private enterprise Soft-Pilot 2000

C++/C# programming, software design.

Laboratory assistant

2003-2005, Moscow State University (Sevastopol Branch)

UNIX system administration and computer laboratory maintenance as a part-time job while studying at the same university.

Publications

  • Aleksandr Drozd, Anna Gladkova, Satoshi Matsuoka. Word Embeddings, Analogies, and Machine Learning: Beyond King - Man + Woman = Queen. Proceedings of COLING 2016, the 26th International Conference on Computational Linguistics: Technical Papers, pages 3519–3530, Osaka, Japan, December 11-17 2016 [pdf]
  • Aleksandr Drozd, Olaf Witkowski, Satoshi Matsuoka and Takashi Ikegami. Critical Mass in the Emergence of Collective Intelligence: a Parallelized Simulation of Swarms in Noisy Environments. Artificial Life and Robotics 2016, volume 21, number 3, pp 317-323 [bib]
  • Mateusz Bysiek, Aleksandr Drozd and Satoshi Matsuoka. Migrating Legacy Fortran to Python While Retaining Fortran-Level Performance through Transpilation and Type Hints. Proceeding of PyHPC '16: the 6th Workshop on Python for High-Performance and Scientific Computing. pp 9-18. [pdf]
  • Anna Gladkova and Aleksandr Drozd. Intrinsic Evaluations of Word Embeddngs: What Can We Do Better? in Proceedings of The 1st Workshop on Evaluating Vector Space Representations for NLP, Berlin, Germany, 2016, pp. 36–42. pdf
  • Annd Gladkova, Aleksandr Drozd and Satoshi Matsuoka. Analogy-based Detection of Morphological and Semantic Relations With Word Embeddings: What Works and What Doesn’t. Proceedings of NAACL-HLT-SRW 2016, pp 8--15. [bib]
  • Aleksandr Drozd, Anna Gladkova and Satoshi Matsuoka Discovering Aspectual Classes of Russian Verbs in Untagged Large Corpora The 2015 IEEE International Conference on Data Science and Data Intensive Systems (DSDIS 2015), pp 61-68, Sydney, Australia, Dec 2015. [bib]
  • Aleksandr Drozd, Anna Gladkova and Satoshi Matsuoka Python, Performance and Natural Language Processing 5th Workshop on Python for High-Performance and Scientific Computing, at Austin, Texas, USA, Nov 2015 in conjunction with SC15. [bib]
  • Aleksandr Drozd, Olaf Witkowski, Satoshi Matsuoka and Takashi Ikegami Signal-Driven Swarming: A Parallel Implementation of Evolved Autonomous Agents to Perform A Foraging Task Proceedings of SWARM 2015 - The First International Symposium on Swarm Behavior and Bio-Inspired Robotics, Kyoto, Oct 2015. [bib]
  • Aleksandr Drozd and Satoshi Matsuoka. HPC and Interactive Big Data Analytics: Case Study of Distributional Semantics. Proceedings of IPSJ SIG Technical Reports 2014-HPC-146, Naha, Oct 2014. [bib]
  • Hideyuki Shamoto, Koichi Shirahata, Aleksandr Drozd, Hitoshi Sato, Satoshi Matsuoka. Large-scale Distributed Sorting for GPU-based Heterogeneous Supercomputers.  Proceedings of 2014 IEEE Conference of Big Data, October 2014, pp 510 - 518. [bib]
  • Aleksandr Drozd, Miquel Pericàs, Satoshi Matsuoka. Efficient String Sorting on Multi- and Many-Core Architectures in Proceedings of IEEE 3rd International Congress on Big Data (2014), pp 637 - 644. [bib]
  • Aleksandr Drozd, Naoya Maruyama, Satoshi Matsuoka. Sequence Alignment on Massively Parallel Heterogeneous Systems in Proceedings of IEEE 26th International Parallel and Distributed Processing Symposium Workshops & PhD Forum (2012), pages 2498 - 2501, ISBN 978-1-4673-0974-5 [bib]
  • Aleksandr Drozd, Naoya Maruyama, Satoshi Matsuoka. A Multi GPU Read Alignment Algorithm with Model-based Performance Optimization, Springer's Lecture Notes in Computer Science N7851 (2012), pages 270-277. [bib]
  • Aleksandr Drozd, Naoya Maruyama, Satoshi Matsuoka. Fast GPU Read Alignment with Burrows Wheeler Transform Based Index, In Companion Proceeding of SC'11 Conference on High Performance Computing Networking, Storage and Analysis, 2011, Pages 21-22 . [bib]
  • Aleksandr Drozd, Naoya Maruyama, Satoshi Matsuoka. Fast Read Alignment with Burrows Wheeler Transform: the GPU Perspective, In Proceedings of the 24th Summer United Workshops on Parallel, Distributed, and Cooperative Processing (SWoPP 2011) , August 2011. [bib]
  • Anna Gladkova and Aleksandr Drozd. Towards Easier Querying of XML-based Linguistic Corpora, Taurida Bulletin of Mathematics and Informatics. #2, 2009, pages 71-77
  • [bib]

Interests

Music

Cello and singing

Photography

Please take a look at nightwind.in

Sports

Hiking and mountaineering, karate