Proceedings of Reisensburg 2013

The conference ‘Statistical Computing’ is held annually at the Reisensburg castle (Germany) by the working group ‘Biostatistics’ of the German Classification Society and by the working group ‘Statistical Computing’ of the German Region of the International Biometric Society and of the German Society of Medical Informatics, Biometry and Epidemiology (GMDS). The conference covers recent topics in Biostatistics and Bioinformatics, with a special focus on applications regarding the computational aspects of these fields of research. This special issue contains selected contributions to the 46th Statistical Computing conference, which took place at Reisensburg castle from 20 to 23 July 2013. All articles underwent the regular reviewing process of Journal of Statistical Computation and Simulation.

Fürstberger et al. [1] propose an algorithm for biological sequence analysis that solves the problem of input data wild cards, which cover missing parts or uncertainty in DNA sequences. The algorithm makes sequence alignment possible even when data are corrupted or incomplete.

Kotthaus et al. [2] provide an analysis of runtime and memory consumption for machine learning programs in the popular statistical programming language R. The paper provides critical insights into runtime problems when handling computation-intensive algorithms in R.

Kraus et al. [3] propose a new algorithm for exhaustive k-nearest-neighbour subspace clustering, which can be applied for small- or medium-sized data sets of approximately 30 features. Based on robustness analysis, the algorithm is able to identify stable candidate subspace clusters.

Krey et al. [4] deal with the topic of cluster analysis in large electrical transmission networks. The authors use cluster algorithms to build models for securing uninterrupted operation of these networks, which are useful for monitoring and protection. Of interest is also that the list of authors includes the father of world football champion Mario Götze.

Lang et al. [5] focus on model selection in survival analysis for high-dimensional data. The authors propose to employ a set of algorithm configuration techniques like F-racing, which are suitable for an automated selection of hyperparameters and the choice between different classes of algorithms in combination with variable selection.