• Mathematical statistics: nonparametric and probability models for ranking data, group-theoretic methods in statistics, rank test statistics, hypotheses testing, ranking and selection procedures, nonparametric density estimation
• Applied statistical modelling: Nonlinear regression, Statistical modeling, Statistical methods in educational and psychological research
• Statistical/Mathematical Consulting:Consultations attended by scientists from various academic institutions (engineering, physics, chemistry, biology, medicine, psychology, education, etc.)

My main research interests lie in mathematical statistics. In particular I am interested in nonparametric and probability models for ranking data, group-theoretic methods in statistics, rank test statistics, hypotheses testing, ranking and selection procedures, nonparametric density estimation.

My early research has been focused on metric approach for analysing partially ranked data. I have completed my Ph.D. thesis entitled "Statistical analysis of partially ranking data" in 1993. The thesis is concerned with ranking and selection problems via decision theoretic approach. The methods include constructing of loss functions using metrics for partially ranked data.

From November 1995 to October 1996, I carried out my research as a postdoctoral fellow in the University of Exeter, UK, with a fellowship awarded by the Royal Society, London. I studied metrics on permutations and started working on rank tests induced by metrics on permutations. I considered a general algebraic structure in which rank statistics are induced by metrics on the permutation group or coset spaces of the permutation group. A class of nonparametric tests based on Chebyshev's metric on permutations is constructed. The exact and asymptotic behaviour of the two sample rank test statistic induced by Chebyshev's metric is studied under the null and some alternative hypothesis.

Some test statistics induced by Chebyshev's metric are related to the well-known statistics based on exceeding observations and the precedence test statistics. Further results concern the power of two sample rank tests based on exceeding observations against location shift alternative,in collaboration with G. Boshnakov, and against Lehmann alternative.

A nonparametric estimation of density functions is another recent topic of my interest. Several publications present results for density estimation for the case when the data are incomplete due to interval censoring.

In 2010 I defended Doctor of Science degree with Habilitation dissertation entitled Metric Methods in Nonparametric Statistical Problems. It summarises all my theoretical results on the topic and include some application.

My current work focusses on the exceedance type tests, which are related to precedence tests. These tests arise as generalization of the two sample statistic induced by Chebyshev's metric. The topic covers three papers in collaboration with N. Balakrishnan.

Go top

In terms of my interests in applied statistics, my research has been explicitly tied to particular substantive questions of researchers from other fields.

My interests in nonlinear regression matched well with that of the mechanical engineering group in Weimar University, Germany. It started in 2003 and was supported by several research grants such as DFG and DAAD. The research was carried out on the problems of statistical modeling of geotechnical processes. The overall objective is to study the hydraulic properties of soils, including statistical modeling of the relationship between unsaturated coeficient of permeability and volumetric water content. Current technology to determine the soil{water characteristic curve (SWCC), and, especially, the unsaturated hydraulic conductivity function is still cumbersome, time consuming and inaccurate for many potential geotechnical and geoenvironmental applications. Some of the results are published in a series of papers.

My applied research also involves several other collaborative basic research studies in the area of mechanics. My current research in biostatistics is within an international group of geneticists studying the genetic variability among bread wheat germplasm. I am responsible for the experimental design, modelling and data analysis. Further research in this direction is in progress.

During the last decade I have consulted several medical doctors on correct application of statistical techniques for analysing medical data. This has been acknowledged in a number of papers and dissertations. In 2005/2008 I served as official scientific advisor of a Ph.D. student from the Sofia Medical University on the statistical aspects of the design and statistical analysis of his (scientific) research.

My interests in psychometrics were motivated by my participation in several national projects on educational testing. My research is primarily on estimation test reliability, meeting model assumptions, estimation measurement error and similar problems. The results are published in the book "Measurement quality of tests", a series of papers in pedagogical journals, and couple of project reports with psychometric analyses. The book is well recognised by Bulgarian test developers and is successfully used as a textbook in several undergraduate and graduate programs. Some of my earlier mathematical papers cover theoretical models for ranking