Online Learning Algorithms For Sequence Prediction, Importance Weighted Classification, And Active Learning
This thesis studies three problems in online learning. For all the problems the proposed solutions are simple yet non-trivial adaptations of existing online machine learning algorithms. For the task of sequential prediction, a modified multiplicative update algorithm that produces small and accurate models is proposed. This algorithm makes no assumption about the complexity of the source that produces the given sequence. For the task of online learning when examples have varying importances, the proposed algorithm is a version of gradient descent in continuous time. Finally, for the task of efficient online active learning, the implementation we provide makes use of many shortcuts. These include replacing a batch learning algorithm with an online one, as well as a creative use of the aforementioned continuous time gradient descent to compute the desirability of asking for the label of a given example. As this thesis shows, online machine learning algorithms can be easily adapted to many new problems.
machine learning; online learning; active learning
Kozen, Dexter Campbell
Hooker, Giles J.; Joachims, Thorsten; Kleinberg, Robert David
Ph. D., Computer Science
Doctor of Philosophy
dissertation or thesis
Showing items related by title, author, creator and subject.
Sullivan, Bilian Ni; Tang, Yi; Marquis, Christopher (2014-08-12)We integrate and extend organizational imprinting and organizational learning theories in a study of how firms’ small-world networks at founding have enduring effects on firm learning. We show that firms embedded in networks ...
Martin, Peter (2002-07)Slides that accompanied the keynote speech delivered at the SubTech 2002 Conference, 3-5 July, University of Warwick
Martin, Peter (2002-08-16)Having done digital course materials from our earliest days by 1995 Tom Bruce and I had added a full on-line law course to the LII's product mix. In the years since, as we have continued, elaborated, and expanded our ...