Concentration of Measure for the Analysis of Randomized Algorithms

Randomized algorithms have become a central part of the algorithms curriculum, based on their increasingly widespread use in modern applications. This book presents a coherent and unified treatment of probabilistic techniques for obtaining high probability estimates on the performance of randomized algorithms. It covers the basic toolkit from the Chernoff–Hoeffding bounds to more sophisticated techniques like martingales and isoperimetric inequalities, as well as some recent developments like Talagrand's inequality, transportation cost inequalities and log-Sobolev inequalities. Along the way, variations on the basic theme are examined, such as Chernoff–Hoeffding bounds in dependent settings. The authors emphasise comparative study of the different methods, highlighting respective strengths and weaknesses in concrete example applications. The exposition is tailored to discrete settings sufficient for the analysis of algorithms, avoiding unnecessary measure-theoretic details, thus making the book accessible to computer scientists as well as probabilists and discrete mathematicians.

1. Chernoff–Hoeffding bounds; 2. Applying the CH-bounds; 3. CH-bounds with dependencies; 4. Interlude: probabilistic recurrences; 5. Martingales and the MOBD; 6. The MOBD in action; 7. Averaged bounded difference; 8. The method of bounded variances; 9. Interlude: the infamous upper tail; 10. Isoperimetric inequalities and concentration; 11. Talagrand inequality; 12. Transportation cost and concentration; 13. Transportation cost and Talagrand's inequality; 14. Log–Sobolev inequalities; Appendix A. Summary of the most useful bounds.

Randomized algorithms have become a central part of the algorithms curriculum, based on their increasingly widespread use in modern applications. This book presents a coherent and unified account of classical and more advanced techniques for analyzing the performance of such algorithms. The presentation emphasizes discrete settings and elementary notions of probability, making it accessible to computer scientists and applied discrete mathematicians.

Devdatt P. Dubhashi is Professor in the Department of Computer Science and Engineering at Chalmers University, Sweden. He earned a Ph.D. in computer science from Cornell University and held positions at the Max-Planck-Institute for Computer Science in Saarbruecken, BRICS, the University of Aarhus and IIT Delhi. Dubhashi has published widely at international conferences and in journals, including many special issues dedicated to best contributions. His research interests span the range from combinatorics to probabilistic analysis of algorithms, and more recently, to computational systems biology and distributed information systems such as the Web.
Alessandro Panconesi is Professor of Computer Science at Sapienza University of Rome. He earned a Ph.D. in computer science from Cornell University and is the recipient of the 1992 ACM Danny Lewin Award. Panconesi has published more than 50 papers in international journals and selective conference proceedings and he is the associate editor of the Journal of Discrete Algorithms and the director of BiCi, the Bertinoro International Center of Informatics. His research spans areas of algorithmic research as diverse as randomized algorithms, distributed computing, complexity theory, experimental algorithmics, wireless networking and Web information retrieval.