|
This book presents theoretical and practical results of information theoretic methods
used in the context of statistical learning. Its major goal is to advocate and promote
the importance and usefulness of information theoretic concepts for understanding
and developing the sophisticated machine learning methods necessary not only to
cope with the challenges of modern data analysis but also to gain further insights
into their theoretical foundations. Here Statistical Learning is loosely defined as a
synonym, for, e.g., Applied Statistics, Artificial Intelligence or Machine Learning.
Over the last decades, many approaches and algorithms have been suggested in the
fields mentioned above, for which information theoretic concepts constitute core
ingredients. For this reason we present a selected collection of some of the finest
concepts and applications thereof from the perspective of information theory as the
underlying guiding principles.We consider such a perspective as very insightful and
expect an even greater appreciation for this perspective over the next years.
The book is intended for interdisciplinary use, ranging from Applied Statistics,
Artificial Intelligence, Applied Discrete Mathematics, Computer Science, Information
Theory, Machine Learning to Physics. In addition, people working in the
hybrid fields of Bioinformatics, Biostatistics, Computational Biology, Computational
Linguistics, Medical Bioinformatics, Neuroinformatics orWeb Mining might
profit tremendously from the presented results because these data-driven areas are
in permanent need of new approaches to cope with the increasing flood of highdimensional,
noisy data that possess seemingly never ending challenges for their
analysis. |