There is a signup schedule for this event (UT EID required).

Typ

e of Talk: UTCS Colloquium/AI

Speaker/Affiliation: Oliver Obst/CSI

RO ICT Centre

Date/Time: Monday April 28 2008 10:30 a.m.

Location: ACES 2.402

Host: Peter Stone

Talk Title: Using Ec

ho State Networks for Anomaly Detection in Underground Coal Mines

Ta

lk Abstract:
We investigated the problem of identifying anomalies in mon

itoring critical gas
concentrations using a sensor network in an undergr

ound coal mine. In this
domain one of the main problems is a provision

of mine specific anomaly
detection with cyclical (moving) instead of fl

atline (static) alarm threshold
levels. An additional practical difficul

ty in modelling a specific mine is the lack
of fully labeled data of no

rmal and abnormal situations. In my talk I''ll present
an approach addr

essing these difficulties based on echo state networks
learning mine spe

cific anomalies when only normal data is available. Echo
state networks

utilize incremental updates driven by new sensor readings
thus enablin

g a detection of anomalies at any time during the sensor network
operati

on. We evaluate this approach against a benchmark -- Bayesian
network ba

sed anomaly detection and observe that the quality of the overall
predi

ctions is comparable to the benchmark. However the echo state
networks

maintain the same level of predictive accuracy for data from multiple
so

urces. Therefore the ability of echo state networks to model dynamical

systems make this approach more suitable for anomaly detection and
predi

ctions in sensor networks.

(This is joint work with X. Rosalind Wang
and Mikhail Prokopenko.)

Speaker Bio:
Oliver Obst obtained a Dip

loma in Computer Science (Informatics) at the University of Koblenz German

y in 1999. At the beginning of 2006 he completed his Ph.D. in Artificial

Intelligence at the same University. Before joining the Adaptive Systems Gr

oup at CSIRO ICT Centre Sydney in 2007
he worked as a postdoctoral re

searcher at the Intelligent Systems Department
University of Bremen Ge

rmany and at the Interdisciplinary Machine Learning
Research Group Uni

versity of Newcastle Australia. His research interest
include simulatio

n prediction and anomaly detection.