Konstantin Vorontsov

PhD (candidate of physics and mathematics);
researcher, Computing Center of the Russian Academy of Sciences (CC RAS), department of computational methods of forecasting;
assistant professor, Moscow Institute of Physics and Technology (MIPT), Department of Control/Management and Applied Mathematics (CMAM);
deputy head of chair «Intelligent systems» CMAM MIPT;
assistant professor, Moscow State University, Faculty of Computational Mathematics and Cybernetics (VMK Faculty);
deputy director for research, FORECSYS.

My main home page on wiki-site MachineLearning.ru (in Russian).

Research interest

Machine learning, data mining, applied statistics, combinatorics.
Inside machine learning: computational learning theory, ensembles, collaborative filtering.

Curriculum vit&ae

1988–1994	MIPT student.
1993–1997	Programmer, RTIAC (research and applied toxicology centre of federal agency on health and social development).
1994–1997	CC RAS post-graduate student.
1996–2000	The development of the programming language ASDIEL for description of learnable algorithms compositions. The development of libraries of classification, clustering and multidimensional scaling algorithms for ASDIEL.
1996–2005	LaTeX-2e in examples. The collection of visual examples for learn LaTeX2e faster (in Russian). My students master LaTeX in two hours!
1997–2001	Expert, Moscow Interbank Currency Exchange (MICEX). The development and supporting of market analysis and surveillance systems SAFRAN and SMARTS.
1998–2006	The development of C++ library of scientific and business charts ChartLib (docs in Russian only).
1999	PhD thesis «The local bases in algebraical approach to recognition problem» (in Russian, see some English papers in publications). Please, do not be confused by our (Russian-translated) terminology; really this work is about ensemble learning, somewhat like boosting, but more general. In particular, nonlinear monotone fusion functions are proposed both for classification and regression.
2000–2002	The development of the C++ library MoTor for Trading Simulation. The unique characteristic of the MoTor is that we model each trading participant (trader, firm of account) separately. The confidential participant data are used to adjust the model. After the model is built all confidential information can be forgotten thus enabling everyone to use and study the model. The development and the project management of the FORECSYS Trading simulator ImiTrade.
2001–now	Researcher, Computing Center of the Russian Academy of Sciences (CC RAS). Deputy director for research FORECSYS.
2002	The development and project management of the FORECSYS System for Monitoring MICEX Futures and Options market.
2002	The development of the internal FORECSYS standard «Components of matrix algorithms maCom» for multidimensional data analysis.
2003–2005	The development of the internal FORECSYS standard «CommFace» for faster development of demo- and pilot-projects.
2003	The development of the pilot version of the FORECSYS Sales Forecasting system Goods4Cast.
2004	The development of the pilot version of the FORECSYS Credit Scoring system ScoringPilot.
2004–now	Assistant professor, deputy head of chair «Intelligent systems» of CMAM MIPT.
2004	The development of one-year course on Machine Learning for MIPT. Lecture notes are here (in Russian).
2004	The project management on Portfolio Selection and Mechanical Trading System.
2005	The project management on Collaborative Filtering and Internet Similarity Map Visualization.
2007	The editing of the proceedings of Russian conference «Mathematical methods of pattern recognition», MMRO-2007. The development of the conference LaTeX style (like a combine style, supporting reviewer notes inside the text).
2007	The development of one-semester course on Applied Statistics for VMK Faculty.
2007–2008	The development of one-year course on Computational Learning Theory for VMK Faculty.
2008–now	The development of the concept, administration, editing the wiki-resource www.MachineLearning.ru. My participant page — хЮБУФОЙЛ:Vokov (in Russian).