Special invited speakers
| Professor Jesper Møller |
| Associate professor Rasmus Waagepetersen |
| Department of Mathematical Sciences, Aalborg University, Denmark |
| Title: Modern Spatial Point Process Modelling and Inference |
|
Abstract: We summarize and discuss the current state of spatial point process theory and directions for future research, making an analogy with generalised linear models and random effect models, and illustrating the theory with various examples of applications. In particular, we consider Poisson, Gibbs, and Cox process models, diagnostic tools and model checking, Markov chain Monte Carlo algorithms, computational methods for likelihood-based inference, and quick non-likelihood approaches to inference. |
| Short CV: |
| Jesper Møller is professor and Rasmus P. Waagepetersen is associate professor at Department of Mathematical Sciences, Aalborg University. They have collaborated for many years on central problems in spatial and computational statistics, where they have a particular interest in theoretical advances in simulation-based inference for spatial point processes and their applications in astronomy, forest ecology, and assessment of wildlife abundance. Their recent monograph is entitled "Statistical inference and simulation for spatial point processes" (Chapman & Hall/CRC). Current research interests of Jesper Møller also includes stochastic geometry and Markov chain Monte Carlo methods, particularly perfect simulation, while Rasmus P. Waagepetersen also is interested in generalized linear mixed models and models for variance heterogeneity in quantitative genetics. |
| Web page: http://www.math.aau.dk/~jm |
|
Web page: http://www.math.aau.dk/~rw
|
| Professor Terry Speed |
| University of California, Berkeley, USA |
| Title: Statistical issues in determining cis-regulatory modules of transcription factors |
|
Abstract: The lectures will outline some of the statistical methods used and challenges ahead as we try to help identify cis-regulatory modules. Loosely speaking, these involve the analysis of a range of types of data, both separately and jointly, with the aim of identifyng individual and later sets of DNA sequence motifs which are involved in the regulation of the expression of particular genes in given tissues under given conditions. The types of data involved include genome sequence data from the organism of interest, and perhaps from related organisms, gene expression data, typically from microarrays, data on the likely location of transcription factor binding sites, typically from chromatin immunoprecipitation (ChIP) assays, with subsequent sequencing or hybridizing to microarrays, and perhaps gene disruption assays. The statistical analyses currently range from motif search algorithms, linear and nonlinear regression with thousands of variables, change-point like algorithms that seek rare events along DNA sequences, methods for finding association between two or more point processes, and more. |
| Short CV: |
| Terry Speed splits his time roughly 50:50 between the Department of Statistics at the University of California, Berkeley in the USA and The Walter and Eliza Hall Institute of Medical Research (WEHI) in Australia each year. His current research concerns the application of statistics to problems in genetics and molecular biology. These have provided many novel challenges of both an applied and a theoretical nature. |
|
Web page: http://www.stat.berkeley.edu/users/terry
|
Invited speakers
| Professor Anders Skrondal |
| Department of Statistics, London School of Economics, London, UK. |
| Title: Some recent developments in latent variable modelling |
|
Abstract: Latent or unobserved variables are commonly used in modern statistical modelling. Examples include random effects in hierarchical or multilevel modelling, common factors in factor analysis, frailties in survival analysis and discrete latent variables in latent class and mixture modelling. Recently, general frameworks have been proposed that take a unified view of these different kinds of latent variable models. The frameworks can handle response variables of different types such as continuous, dichotomous, ordinal, nominal as well as counts and survival. Different types of responses can then be combined to produce for instance joint models of survival and dropout. Another useful feature of the frameworks is that latent variables may be continuous (parametric or non-parametric) or discrete or mixed continuous-discrete. Challenges include estimation and inference for complex models. The challenges become compounded when complex sampling designs are used. |
| Short CV: |
|
Anders Skrondal is Professor of Statistics and Chair in Social Statistics, Department of Statistics, London School of Economics. He was previously Head of the Biostatistics Group at the Division of Epidemiology, Norwegian Institute of Public Health.
Skrondal's research interests include topics in biostatistics, social statistics and psychometrics. His Dr.Philos dissertation in biostatistics was awarded the "1997 Psychometric Society Dissertation Prize". Recently, he has concentrated on collaborative work on the development of the Generalized Linear Latent and Mixed Model (GLLAMM) framework. Outcomes of this project include numerous papers published in (bio)statistical, psychometric and econometric journals as well as two recent books; "Generalized Latent Variable Modeling" published by Chapman & Hall/CRC in 2004 and "Multilevel and Longitudinal Modeling using Stata" published by Stata Press in 2005. |
|
Web page: http://www.gllamm.org/anders.html
|
| Professor Yudi Pawitan |
| Department of Medical Epidemiology and Biostatistics, Karolinska Institutet, Stockholm, Sweden |
| Title: Multidimensional local false discovery rate |
|
Abstract: The false discovery rate (fdr) is a key tool for statistical assessment of differential expression (DE) in microarray studies. It is, however, well known that overall control of the fdr alone is not sufficient to address the problem of genes with small variance, which suffer from a disproportional high rate of false positives. Graphical tools and modified test statistics have been proposed for dealing with this problem, but there is currently no procedure for controlling the fdr directly. Methods: We generalize the local fdr called fdr2d - as a function of multiple statistics, combining a common test statistic for assessing differential expression with standard error information. Results: The fdr2d allows an objective assessment of differential expression as a function of gene variability. Furthermore, the fdr2d has comparable performance to other methods that model the variance explicitly or to the theoretically optimal procedure. |
| Short CV: |
| Yudi Pawitan graduated from the University of California at Davis in 1987 and had worked in University of Washington and National University of Ireland. He has been at Karolinska Institutet since 2001, and currently interested in applied and methodological problems in statistical genetics and high-throughput genomic data. |
|
Web page: http://www.mep.ki.se/~yudpaw
|
| Professor Tobias Rydén |
| Centre for Mathematical Sciences, Lund University, Lund, Sweden |
| Title: Hidden Markov and state space models - from likelihood theory to computational statistics |
|
Abstract: During the the last 10-15 years, hidden Markov and more general state space models have undergone a rapid development in terms of theoretical and computational statistics. On the inferential side most of the focus has been on likelihood theory, which in many respects today is rather complete -- at least for models with compact state space. However, except for models with finite state space and linear Gaussian models, the likelihood cannot be computed, let alone maximised, exactly. Many procedures to compute approximations have been proposed, with so-called sequential Monte Carlo methods, or particle filters, having received much attention recently. The topic of this talk is the merge of these two areas: finding numerical methods that produce approximate ML estimators that at least asymptotically have favourable properties. We will talk about estimators derived from numerical approximations of the likelihood function, Monte Carlo EM algorithms, simulated annealing-type algorithms, and also about their performance in terms of consistency and efficiency. |
| Short CV: |
| Tobias Rydén received a PhD from Lund University in 1993. After a period as postdoc at UC Berkeley's statistics department,
he has been based in Lund, doing longer or shorter research visits
to the universities of Glasgow, Cambridge and ENSAE and ENST (Paris).
His general areas of interest are inference in stochastic processes and in missing data models, both from a theoretical and a computational perspective. He has some 20 journal publications on the topics of hidden Markov and state space models, much of which is summarised in the recent monograph with Olivier Cappé and Eric Moulines, Inference in Hidden Markov Models (Springer, 2005). |
|
Web page: http://www.maths.lth.se/matstat/staff/tobias
|
| Professor Fred Espen Benth |
| Centre of Mathematics for Applications, Department of Mathematics, University of Oslo, Norway |
| Title: Mathematical finance for energy markets: stochastic models and pricing of derivatives |
|
Abstract: We discuss different mean-reversion stochastic processes with jumps for modelling the evolution of temperature and spot prices of energies. The typical examples we have in mind are the weather derivatives market and the liberalized electricity and gas markets. Based on these models, we derive prices for different energy forward/futures contracts which are settled over a time period rather than at a fixed settlement time. Forward/futures contracts on temperature is usually written on some averaging over warm or cold days, and we derive explicit formulas based on a temperature model with seasonal variance. We further introduce the Heath-Jarrow-Morton (HJM) approach to energy forward/futures modelling, an approach inspired from fixed-income markets. A special emphasis is put on modelling a reasonable term-structure for the volatility of forward/futures prices. Different methods to price options are analyzed. |
| Short CV: |
| Fred Espen Benth is a professor of mathematical finance at the Centre of Mathematics for Applications (CMA), University of Oslo, Norway. His research interests include pricing and hedging in incomplete markets and stochastic portfolio optimization, where he has several publications. Outside academia, Benth holds a part-time position as an advisor for a major life insurance company in Norway, where he is leading a project in pricing interest rate guarantees. |
|
Web page: http://www.math.uio.no/~fredb
|
| Professor Rasmus Nielsen |
| Centre for Bioinformatics, University of Copenhagen, Denmark |
| Title: Statistical inference in population genetics |
|
Abstract: Population genetics is the study of genetic variability within and between populations. It has recently gained new importance as population genetic methods are used in many areas of medical research, particularly in association mapping studies. However, inference in population genetic models present special computational and statistical problems. Most models are based on well-described stochastic process theory, but likelihood functions can usually only be calculated using computationally slow Monte Carlo procedures. The fundamental problem is that samples from different individuals are not iid but are correlated through an underlying genealogical tree. Even in some of the most simple models the likelihood function cannot be calculated without the use of very slow MCMC or Sequential Importance Sampling methods - and for many problems even these methods do not work for realistic sized samples. Researchers must often instead rely on Ad hoc inference procedures with poor or unknown statistical properties. In this talk I will give an introduction to some of the statistical problems encountered in population genetics and discuss some of the solutions that have been proposed. |
| Short CV: |
| Rasmus Nielsen has spent the last 10 years in the United States at UC-Berkeley, Harvard University and Cornell University. Last year he returned to Denmark in an Ole Rømer fellowship awarded by the Danish national Science Foundation and was subsequently offered a full professorship in the Department of Biology. RN's research focuses on statistical methods in molecular evolution and population genetics. He is an associate editor of the Journal of Molecular Evolution, has served a guest editor of Systematic Biology and has been the editor of a new book entitled Statistical Methods in Molecular Evolution. He has published 83 scientific papers (69 peer reviewed papers and 11 invited contributions and book chapters) including 6 papers in Nature and Science since 1997. In addition to various Danish sources, RN has received grant support from the National Science Foundation (NSF), the National Institute of Health (NIH) and the Human Frontier in Science Program (HFSP). He has also received fellowships from the Fulbright Commission and the Sloan Research Foundation. He is a frequent invited speaker at international meetings and symposia and has given more than 50 invited talks. He has served as a grant reviewer for various granting agencies in 5 different countries including the BBSRC (UK) and the NSF (US), as an appointed member of the Regional Advisory Board of the International Biometrics Society, Eastern North America Region, and as a member of the Data Analysis Working Group of the International DNA Barcoding Consortium. He has been on numerous administrative committees at Cornell University including the Curriculum Committee of the College of Agriculture and Life Sciences and the Genomics Task Force. He has taught classes in Molecular Evolution, Statistical Genomics, and Bioinformatics. |
|
Web page: http://www.binf.ku.dk/users/rasmus/webpage/ras.html
|
| Professor Olle Häggström |
| Chalmers University of Technology, Gothenburg, Sweden |
| Title: Problem solving is often a matter of cooking up an appropriate Markov chain |
|
Abstract: By means of a series of examples, taken from classic contributions to probability theory as well as from my own practice, I will try to convince the audience of the claim made in the title of the talk. Along the way, I will have reason to discuss topics such as coupling, correlation inequalities, and percolation. |
| Short CV: |
| Olle Häggström was born in 1967 and completed his Ph.D. in mathematical statistics in 1994 at Chalmers University of Technology, where he is now a professor. His main research interests are in discrete probability and problems at the interface between probability and statistical mechanics. He is chairman of the Swedish Mathematical Society, and a member of the Royal Swedish Academy of Sciences. |
|
Web page: http://www.math.chalmers.se/~olleh
|
| David Spiegelhalter |
| MRC Biostatistics Unit, Institute of Public Health, Cambridge, United Kingdom |
| Title: Monitoring performance in the UK health-care system: the role of statistical methods |
|
Abstract: |
| Short CV: |
| David Spiegelhalter FRS is a Senior Scientist at the MRC Biostatistics Unit. He has worked extensively on the theory and applications of Bayesian methods, including the BUGS project. He gave evidence to both the Bristol and Shipman Inquiries, and is now a statistical consultant to the Healthcare Commission. |
|
Web page: http://www.mrc-bsu.cam.ac.uk/BSUsite/AboutUs/People/davids.shtml
|
Sessions
| Professor Eva B. Vedel Jensen |
| Department of Mathematical Sciences, University of Aarhus, Denmark |
| Short CV: |
| Eva B. Vedel Jensen is professor at Department of Mathematical Sciences, University of Aarhus. She has for many years worked in the research field of stochastic geometry and stereology. Together with Adrian Baddeley, University of Western Australia, she published in 2004 the monograph entitled "Stereology for Statisticians" (Chapman & Hall/CRC). In recent years she has widened her research interests which now also include spatio-temporal modelling and its applications in neuroscience. Since 2004, she has been the scientific director of the Thiele Centre for Applied Mathematics in Natural Science, University of Aarhus, see http://www.thiele.au.dk |
|
Web page: http://home.imf.au.dk/eva
|
| Associate professor Anders Rahbek |
| Department of Applied Mathematics and Statistics, Institute for Mathematical Sciences, University of Copenhagen, Denmark |
| Short CV: |
|
Research Interests: Linear and Nonlinear (Multivariate) Time series analysis, including cointegration analysis in Econometrics and financial time series modelling in Financial Econometrics.
Recent publications in journals: |
|
Web page: http://www.math.ku.dk/~rahbek
|
| Professor Mats Rudemo |
| Mathematical Statistics, Chalmers University of Technology, Gothenburg, Sweden |
| Short CV: |
| Mats Rudemo is professor in mathematical statistics at the Royal Veterinary and Agricultural University, Copenhagen, Denmark. Since 1997 he is also director of Stochastic Centre, Gothenburg. His research interests include statistics for microarrays and 2D electrophoresis, image analysis, spatial statistics, particle tracking and precision forestry. |
|
Web page: http://www.math.chalmers.se/~rudemo
|
| Professor Odd O. Aalen |
| Department of Biostatistics, Institute of Basic Medical Sciences, University of Oslo, Norway |
| Short CV: |
| TBA |
| Web page: |
| Professor Ørnulf Borgan |
| Department of Mathematics, University of Oslo, Norway |
| Short CV: |
|
Address, etc: Department of Mathematics University of Oslo P.O.Box 1053 Blindern N-0316 Oslo, Norway e-mail: borgan@math.uio.no Personal facts: Born: 8 April 1950; Citizenship: Norwegian. Accademic degrees: 1984: Dr. philos. (Ph.D.) in statistics, University of Oslo. 1976: M.sc. in statistics, University of Oslo. Positions held: 1993 - present: Professor in statistics, University of Oslo. 1984 - 92: Associate professor in statistics, University of Oslo. 1983: Associate professor in statistics, Agricultural University of Norway. 1980 - 82: Research fellow in statistics, University of Oslo. 1977 - 79: Assistant professor in insurance mathematics, University of Copenhagen. Editorial work: The Annals of Statistics: Associate editor 1998 - 2003. Scandinavian Journal of Statistics: Associate editor 1997 - 2003; editor elect (with Bo Lindqvist): 2007-2009. Research interests: Statistical models and methods based on counting processes; Survival and event history analysis; Case-control designs; Statistical methods in genomics. |
|
Web page: http://folk.uio.no/borgan
|
| Associate professor Inge Henningsen |
| Center for Statistics, Copenhagen Business School, Denmark |
| Short CV: |
|
Inge Henningsen, cand. stat. Lektor ved Afdeling for Anvendt Matematik og Statistik Det naturvidenskabelige fakultet Københavns Universitet |
|
Web page: http://www.stat.ku.dk/~inge
|
| Professor Helle Rootzén |
| Informatics & Mathematical Modelling Statistics section, Technical University of Denmark |
| Short CV: |
| Helle Rootzén is Professor at the Technical University of Denmark. Her research interests include chemometrics, breast cancer statistics, spatial-temporal models for environmental problems, and a range of topics in industrial statistics. She is leader of the project "Continuing education voucher systems: A flexible and targeted statistics programme based on learning objects and blended learning" which is financed by the Ministry of Science, Technology and Innovation. |
|
Web page: http://www2.imm.dtu.dk/~hero
|
| Professor Bent Jørgensen |
| Department of Statistics, University of Southern Denmark |
| Short CV: |
| Bent Jørgensen is Professor of Mathematical Statistics at the University of Southern Denmark. He received a Cand.Scient. degree from the University of Aarhus, 1979, a Ph.D. from Odense University, 1987, and is Dr.Scient. from Aalborg University, 1997, all in the area of statistics. He was Assistant Professor (1980-84) and Associate Prefessor (1984-1987) at Odense University, Associate Professor at the Institute of Pure and Applied Mathematics, Brazil (1987-92), and Associate Professor at the University of British Coprogramme based on learning objects and blended learningprogramme based on learning objects and blended learninlg which is financed by the Ministry of Science, Technology and Innovation.umbia (1992-97). His current research interests include statistical inference, estimating functions, longitudinal data analysis, random-effects models and chemometrics. He is program coordinator for the web-based Master of Applied Statistics. |
|
Web page: http://www.stat.sdu.dk/matstat/bent
|
| Professor Rolf Sundberg |
| Mathematical statistics, Stockholm University, Sweden |
| Short CV: |
| Rolf Sundberg, professor of mathematical statistics at Stockholm University, with a long interest in chemometrics, particularly calibration problems. In the 1994 Nordic conference he was an invited speaker, talking on multivariate calibration. More recently he was the Kowalski prize winner (jointly with Marie Linder), for the best theoretical paper in Journal of Chemometrics 2002-2003. |
|
Web page: http://www.math.su.se/~rolfs
|
| Professor Henrik Madsen |
| Informatics and Mathematical Modelling, Technical University of Denmark, Denmark |
| Short CV: |
| Henrik Madsen received the M.Sc. in Engineering in 1982, and the Ph.D. in Statistics in 1986, both at the Technical University of Denmark. He was appointed ass. prof. in Statistics in 1986, assoc. prof. in 1989, and professor in Statistics with a special focus on Stochastic Dynamic Systems in 1999. He has been external lecturer at a number of universities. He is involved in a large number of cooperative projects with other universities, research organizations and industrial partners. His main research interest is related to analysis and modelling of stochastic dynamics systems. This includes signal processing, time series analysis, identification, estimation, grey-box modelling, prediction, optimization and control. The applications are mostly related to Energy Systems, Informatics, Environmental Systems, Bioinformatics, Process Modelling and Finance. He has authored or co-authored approximately 280 papers and technical reports, and about 10 educational texts. He is the leader of Center for High Performance Computing at DTU which was opened by the Danish Minister of Research in February 2002. |
|
Web page: http://www2.imm.dtu.dk/~hm
|
| Associate professor Svend Kreiner |
| Department of Biostatistics, University of Copenhagen, Denmark |
| Short CV: |
| TBA |
|
Web page: http://www.biostat.ku.dk/~skm/skm/
|
| Professor Jouko Lampinen |
| Laboratory of Computational Engineering, Helsinki University of Technology, Helsinki, Finland |
| Short CV: |
| Jouko Lampinen is a professor in computational information technology in the helsinki University of Technology. He received M.Sc. degree in physics from University of Kuopio, Finland, in 1988, and Ph.D. in information technology from Lappeenranta University of Technology, Finland, in 1993. His current research interest include machine vision and statistical learning methods, especially Bayesian and MCMC techniques. |
|
Web page: http://www.hut.fi/~jlampine
|