Boris Mirkin

1967-1982   Researcher,   Head of Lab.(1975)   Institute of Economics , Siberian Branch of Soviet Academy   of Sciences, Novosibirsk, Russia 

1982-2000    Senior researcher, Central Economics-Mathematics Institute of RAS, Moscow

2000-2008    Professor,   Birkbeck University of London, London UK

2008-pres..   Professor,     National Research University Higher School of Economics,Moscow RF

Part time

1970-1982    Assoc.Professor        Department of Mathematics-Economics Methods, Novosibirsk,  State University, Novosibirsk

1984-1991    Assoc.Professor        Department of Mathematics and Statistics, Higher School ofTrade Unions of the USSR, Moscow

2008-2010    Professor                   Birkbeck University of London, London UK 

2009-pres.    Leading research      Laboratory on Decision Choice and Analysis at HSE Moscow                                  

• A method for pattern recognition utilizing information of the structure of pattern neighborhood both for a single classifier and multiple classifiers (in collaboration with Zacharias Voulgaris, Birkbeck College Ph.D. Student; published in 2010)
• Experimental assessment of the author-proposed “intelligent” version of the popular K-means clustering method (iK-Means). It is experimentally shown that the iK-means method outperforms seven other methods in various conditions (in collaboration with Ming-Tso Chiang, Birkbeck College Ph.D. Student; published in 2010)
• A method for representation of  research organization’s activities by means of optimal lifting over a taxonomy of the domain under consideration (in collaboration with Prof. T. Fenner, Birkbeck University of London, UK and Dr. S. Nascimento, New University of Lisbon, Portugal, partially published in 2010)
• A system and description of data analysis main methods (published as a textbook, Core Concepts in Data Analysis: Summarization, Correlation, Visualization, Springer, 2011, 388 p.)
• Extending k-means clustering to the case of Minkowski distance with automatic choice of feature weights (jointly with Birkbeck PhD student Renato Amorim, published 2012)
• Method of annotated suffix tree for automation of the analysis of the relationship between a set of key phrases and a collection of unstructured texts applied to problems such as annotation of research papers, completion of taxonomies, etc. (jointly with E. Chernyak, published in 2012-2015, mostly in Russian)
• Method for automatic combining multiple criteria in a convex combination; its verification at specially developed models of data generation  (jointly with NRU HSE PhD student M. Orlov, pubished in 2014-1015)
• Method for direct assessment of the level of research results using a taxonomy of the research domain; its partial verification at the data analysis domain (jointly with NRU HSE PhD student M. Orlov, pubished in 2014-1015)
• Adaptation of my earlier method for anomalous clustering to problems of the analysis of temperature maps of the sea surface (jointly with Dr. S. Nascimento, New University of Lisbon, Portugal, published 2015)
• A novel method for divisive cluster-analysis with random projections of the data (jointly with Е. Kovaleva, published 2015)
• Method for accelerating agglomerative clustering by using my method of anomalous clusters (jointly with R. De Amorin (Essex UK) and V. Makarenkov (Montreal Canada), published in 2016)
• Method for concensus partitioning by using Minkowski metrics (jointly with R. De Amorin (Essex UK) and V. Makarenkov (Montreal Canada), published in 2017)
• Method for finding super-clusters to consist of "tight" core and shell, for datasets represented by similarity graphs both in statics and dynamics (joinyly with I. Rodin, partly published in 2017)
• Method for parsimonious lifting of leaf clusters in the rooted tree of a domain taxonomy to the higher ranks is extended to the case at which the cluster is fuzzy (jointly with D. Frolov,  T. Fenner (London UK) and S. Nascimento (LIsbon Portugal), published in 2019) 
• A hybrid method for the analysis of tendencies in researh within a domain by using a domain taxonomy and a collection of papers within the domain, combining my three earlier developed mehods  (string-to-text relevance scoring, fuzzy clustering, and  parsimonious lifting of fuzzy clusters), jointly with D. Frolov,  T. Fenner (London UK) and S. Nascimento (LIsbon Portugal)  published in 2020

• Method of anomalous cluster in delineating upwellings in coastal waters (jointly with S. Nascimento (New University of Lisbon, Portugal), 2014-2015
• Methods for evaluation of the quality/level of research results (jointly with PhD student M.A. Orlov) 2013-2015
• Methods for stratification of objects over a set of criteria (jointly with PhD student M.A. Orlov) 2012-2015
• Methods for analysis and visualization of unstructured texts (jointly with E.L. Chernyak); supported by grants from the Academic Fund of HSE to lead a group of 5-6 students (2012-2015)
• An additive spectral method for cluster analysis of similarity matrices, its experimental verification and application (in collaboration with Dr. S. Nascimento, New University of Lisbon, Portugal).
• Stopping criteria for divisive hierarchical clustering algorithms (in collabortion with E.Kovaleva, HSE student)
• Approximate tri-clusters (in collabortion with A. Kramarenko, D. Gnatyshak, D.I.Ignatov,
• Methods for interpretation of thematic clusters over hierarchical taxonomies and other cognitive structures (in collaboration with Prof. T. Fenner, Birkbeck University of London, UK, and Dr. S. Nascimento, New University of Lisbon, Portugal)
• Methods for analysis and classification of a corpus of unstructured text documents by using the techniques of annotated suffix trees and taxonomies of specific domains. These methods are to be applied to programmes of teaching courses, taught in HSE (in collaboration with E.Chernyak, O.Chugunova, J.Askarova, HSE students). This work is part of the research project “Teacher–Students: Development and adaptation of cluster analysis methods to automate the analysis of textual information by using a specific domain’s ontology”. Six students of the School of Applied Mathematics and Informatics take part in this project, supported by the HSE Research Foundation 2011-2012. 

Participation in conferences in 2010 – 2011

• 11th Russian national symposium “Strategy planning and enterprises development”, April 2010, CEMI of RAS, invited plenary talk “Representing research organization activities, considering the structure of the research area domain: the cluster-lift method”
• 4th International Conference “Knowledge Science, Engineering and Management”, September 2010, Belfast, Northern Ireland, UK. Paper: Constructing and mapping fuzzy thermatic clusters to higher ranks in a taxonomy (jointly with S. Nascimento, T. Fenner, and L.M. Pereira) has been selected among six others to have been published in International Journal of Software and Informatics, 2010. № 4(3). C. 257—275.
• 13th International Conference on Rough Sets, Data Mining, Fuzzy Sets and Granular Computing (RSFDGrC-2011), Moscow HSE June 2011 – General Chair
• 4th International Conference on Pattern Recognition and Machine Intelligence (PReMI-2011), Moscow HSE June 2011, Invited plenary talk.
• International Conference on Classifications, Glasgow UK, July 2011 – Member of the organizing committee, a paper
• Biannual conference of International Federation of Classification Societies, Frankfurt, Germany, August 2011 – Member of the Program Committee, a paper
• 2011 meeting of the European Mathematical Psychology Group, August 2011, Telecom, Paris, France – Member of the Scientific Committee  

HSE grant for individual research projects 2009: 09-01-0071 "Interrelation between spectral and approximation approaches to cluster analysis"

HSE grant  11-04-0019 “Teacher–Students: Development and adaptation of cluster analysis methods to automate the analysis of textual information by using a specific domain’s ontology”  (Six students of the School of Applied Mathematics and Informatics take part in this project)
HSE Academic Programme grant Research and Study Lab 13-05-0047 Methods for text visualization using suffix trees, multifacet classifications and hierarchic ontologies: algorithms and software (six students involved)