N. Shiryaev, "Quickest detection of the spontaneous appearing
effects - fifty years later"
Abstract: We describe the very beginning
(since 1959) of our work with A.N.Kolmogorov on the radiolocation
problems which led us to the formulation and solution of the different
versions of the quickest detection of the spontaneous appearing
effects. A panorama of these results, their developments with
numerous applications in different fields will be presented.
Dragalin, "Adaptive designs in dose-ranging clinical trials"
Abstract: The overall goal of dose-ranging
studies is to establish the existence, nature and extent of dose
effect. Recently, a great effort has been made to investigate
and develop designs and methods for efficient learning about dose
response and more accurate and faster decision making on dose
selection and improved labeling. The focus is on adaptive and
model-based dose-ranging designs.
In this presentation, I will compare the
performance of the different adaptive designs in learning about
and estimating the dose response relationship. The evaluation
of statistical operational characteristics of these new adaptive
designs has been mainly done through comprehensive simulation
studies. There is a need for new theoretical developments in the
methodology for planning, monitoring and analysis for these sequential
Huková, "Sequential nonparametric estimation"
Abstract: The talk will start with a short
partial survey of results on sequential nonparametric procedures.
Then it will focus on sequential confidence regions in location
models as well as in nonparametric regression setup. Also, semiparametric
sequential test procedures when a training sample is available
will be covered. Both independent and dependent observations will
be considered. Possibility of application of a suitable version
of bootstrap method will be also discussed. Theoretical results
will be accompanied by simulation study.
Moshe Pollak "The Shiryaev-Roberts
Changepoint Detection Procedure in Retrospect - Theory and Practice"
Abstract: A retrospective view of the Shiryaev-Roberts
procedure is presented, placing it in a historical context and
describing its evolution into a powerful changepoint detection