Abstract Details
Activity Number:
|
180
|
Type:
|
Contributed
|
Date/Time:
|
Monday, August 5, 2013 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistical Computing
|
Abstract - #307692 |
Title:
|
Making Rules Human-Interpretable for Alarm Prediction in Sensor Network
|
Author(s):
|
Hongfei Li*+ and Buyue Qian and Dhaivat Parikh and Arun Hampapur
|
Companies:
|
IBM T. J. Watson Research and UC Davis and IBM GBS and IBM Research
|
Keywords:
|
Alarm Prediction ;
Big Data ;
Human Interpretable ;
Sensor Network
|
Abstract:
|
Sensor network is deployed to monitor equipment operations and issue alarms when potential malfunctioning is detected. Once the sensor network triggers the alarms of the most severe category, immediate interruption of equipment operations is required, which costs significantly to the network velocity. Unlike the existing work, we do not aim to improve the accuracy of the current alarm rules that operators deploy. Instead, we provide prediction of those alarms in the most severe category several days in advance so that operators have sufficient time to respond. There are three concerns we bear in mind when developing prediction model. The first is to keep false alarm rate very low due to the operation resource constraints. The second is to develop big data techniques to analyze large volume of sensor data around 2TB collected in half a year. The last is to provide human interpretable rules to facilitate decision process by operations group. This work provides solutions to address these issues.
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2013 program
|
2013 JSM Online Program Home
For information, contact jsm@amstat.org or phone (888) 231-3473.
If you have questions about the Continuing Education program, please contact the Education Department.
The views expressed here are those of the individual authors and not necessarily those of the JSM sponsors, their officers, or their staff.
Copyright © American Statistical Association.