Switching
Patterns in data often change with time. We take a worst-case approach to designing adaptive algorithms to deal with such non-stationarity.
-
A closer look at adaptive regret
Dmitry Adamskiy, Wouter M. Koolen, Alexey Chernov, Vladimir Vovk
Journal of Machine Learning Research, 17(23):1–21, April 2016.
-
Open problem: Shifting experts on easy data
Manfred K. Warmuth, Wouter M. Koolen
In Proceedings of the 27th Annual Conference on Learning Theory (COLT), pages 1295–1298, June 2014.
-
Universal codes from switching strategies
Wouter M. Koolen, Steven de Rooij
IEEE Transactions on Information Theory, 59(11):7168–7185, November 2013.
-
Putting Bayes to sleep
Wouter M. Koolen, Dmitry Adamskiy, Manfred K. Warmuth
In Advances in Neural Information Processing Systems (NeurIPS) 25, pages 135–143, December 2012.
-
A closer look at adaptive regret
Dmitry Adamskiy, Wouter M. Koolen, Alexey Chernov, Vladimir Vovk
In Proceedings of the 23rd International Conference on Algorithmic Learning Theory (ALT), LNAI 7568, pages 290–304. Springer, October 2012.
-
Combining Strategies Efficiently: High-quality Decisions from Conflicting Advice
Wouter M. Koolen
PhD thesis, Institute of Logic, Language and Computation (ILLC), University of Amsterdam, January 2011. cum laude.
-
Switching between hidden Markov models using Fixed Share
Wouter M. Koolen, Tim van Erven
Computing Research Repository (CoRR), abs/1008.4532, February 2010.
-
Freezing and sleeping: Tracking experts that learn by evolving past posteriors
Wouter M. Koolen, Tim van Erven
Computing Research Repository (CoRR), abs/1008.4654, February 2009.
-
Combining expert advice efficiently
Wouter M. Koolen, Steven de Rooij
In Proceedings of the 21st Annual Conference on Learning Theory (COLT), pages 275–286, June 2008.
-
Combining expert advice efficiently
Wouter M. Koolen, Steven de Rooij
Computing Research Repository (CoRR), abs/0802.2015, February 2008.