Chis, Tiberiu S. ;
Harrison, Peter G.
Incremental HMM with an improved BaumWelch Algorithm
Abstract
There is an increasing demand for systems which handle higher density, additional loads as seen in storage workload modelling, where workloads can be characterized online. This paper aims to find a workload model which processes incoming data and then updates its parameters "onthefly." Essentially, this will be an incremental hidden Markov model (IncHMM) with an improved BaumWelch algorithm. Thus, the benefit will be obtaining a parsimonious model which updates its encoded information whenever more real time workload data becomes available. To achieve this model, two new approximations of the BaumWelch algorithm are defined, followed by training our model using discrete time series. This time series is transformed from a large network trace made up of I/O commands, into a partitioned binned trace, and then filtered through a Kmeans clustering algorithm to obtain an observation trace. The IncHMM, together with the observation trace, produces the required parameters to form a discrete Markov arrival process (MAP). Finally, we generate our own data trace (using the IncHMM parameters and a random distribution) and statistically compare it to the raw I/O trace, thus validating our model.
BibTeX  Entry
@InProceedings{chis_et_al:OASIcs:2012:3761,
author = {Tiberiu S. Chis and Peter G. Harrison},
title = {{Incremental HMM with an improved BaumWelch Algorithm}},
booktitle = {2012 Imperial College Computing Student Workshop},
pages = {2934},
series = {OpenAccess Series in Informatics (OASIcs)},
ISBN = {9783939897484},
ISSN = {21906807},
year = {2012},
volume = {28},
editor = {Andrew V. Jones},
publisher = {Schloss DagstuhlLeibnizZentrum fuer Informatik},
address = {Dagstuhl, Germany},
URL = {http://drops.dagstuhl.de/opus/volltexte/2012/3761},
URN = {urn:nbn:de:0030drops37613},
doi = {http://dx.doi.org/10.4230/OASIcs.ICCSW.2012.29},
annote = {Keywords: hidden Markov model, BaumWelch algorithm, Backward algorithm, discrete Markov arrival process, incremental workload model}
}
Keywords: 

hidden Markov model, BaumWelch algorithm, Backward algorithm, discrete Markov arrival process, incremental workload model 
Seminar: 

2012 Imperial College Computing Student Workshop

Issue date: 

2012 
Date of publication: 

2012 