We also introduce new methods of dimension reduction and clustering for time series data and describe some theoretical results on the weighted correlation coefficients in separate sections.
The determination of sample size and the evaluation of power are fundamental and critical elements in the design of clinical trials. If the sample size is too small, important effects may go unnoticed; if the sample size is too large, it represents a waste of resources and unethically puts more participants at risk than necessary. Recently many clinical trials have been designed with more than one endpoint considered as multiple primary or co-primary, creating a need for new approaches to the design and analysis of these clinical trials. The book focuses on the evaluation of power and sample size determination when comparing the effects of two interventions in superiority clinical trials with multiple endpoints. Methods for sample size calculation in clinical trials where the alternative hypothesis is that there are effects on ALL endpoints are discussed in detail. The book also briefly examines trials designed with an alternative hypothesis of an effect on AT LEAST ONE endpoint with a prespecified non-ordering of endpoints.
Since Sir David Cox’s pioneering work in 1972, the proportional hazards model has become the most important model in survival analysis. The success of the Cox model stimulated further studies in semiparametric and nonparametric theories, counting process models, study designs in epidemiology, and the development of many other regression models that could offer more flexible or more suitable approaches in data analysis.Flexible semiparametric regression models are increasingly being used to relate lifetime distributions to time-dependent explanatory variables. Throughout the book, various recent statistical models are developed in close connection with specific data from experimental studies in clinical trials or from observational studies.
The book presents selected papers from the Second European Conference on Data Analysis, held at Jacobs University Bremen in July 2014. This conference unites diverse researchers in the pursuit of a common topic, creating truly unique synergies in the process.
The 15 articles are a representative sample of the 336 contributed papers presented at the conference. They cover topics such as high-dimensional data modelling, inference for stochastic processes and for dependent data, nonparametric and goodness-of-fit testing, nonparametric curve estimation, object-oriented data analysis, and semiparametric inference.
The aim of the ISNPS 2014 conference was to bring together recent advances and trends in several areas of nonparametric statistics in order to facilitate the exchange of research ideas, promote collaboration among researchers from around the globe, and contribute to the further development of the field.