Conjoint Modeling of Temporal Dependencies in Event Streams

Content Provider	Microsoft Research
Author	Parikh, Ankur P. Gunawardana, Asela Meek, Christopher
Copyright Year	2012
Abstract	Many real world applications depend on modeling the temporal dynamics of streams of diverse events, many of which are rare. We introduce a novel model class, ConjointPiecewise-Constant Conditional Intensity Models, and a learning algorithm that together yield a data-driven approach to parameter sharing with the aim of better modeling such event streams. We empirically demonstrate that our approach yields more accurate models of two real world data sets: search query logs and data center system logs.
Language	English
Publisher	UAI Bayesian Modelling Applications Workshop
Publisher Date	2012-08-18
Access Restriction	Open
Rights Holder	Microsoft Corporation
Subject Keyword	Machine learning intelligence
Content Type	Text
Resource Type	Proceeding