Learning and Inference in Queueing Networks
Charles Sutton and Michael I Jordan
In: AISTATS 2010, 13-15 May 2010, Sardinia, Italy.
Probabilistic models of the performance of
computer systems are useful both for predicting system performance in new conditions,
and for diagnosing past performance prob-
lems. The most popular performance models are networks of queues. However, no cur-
rent methods exist for parameter estimation
or inference in networks of queues with missing data. In this paper, we present a novel
viewpoint that combines queueing networks
and graphical models, allowing Markov chain
Monte Carlo to be applied. We demonstrate
the eectiveness of our sampler on real-world
data from a benchmark Web application.