|
In This Issue:
|
March/April 2007
|
 |
|
|
•
|
Charles River Analytics Scientist Serves as Guest Expert at Multi-Sensor Data Fusion Seminar
|
|
|
|
•
|
BNet.EngineKit User Tip: Saving Files
|
|
|
|
|
•
|
Bayesian Networks for Cardiovascular Monitoring
|
|
|
|
•
|
View Back Issues of BNet.News
|
|
|
|
|
•
|
BNet.Builder User Tip: Undo/Redo
|
|
|
|
•
|
Publicize Your Belief Network Success Story
|
|
|
|
WASHINGTON, D.C. – Dr. Subrata Das, Chief Scientist at Charles River Analytics, served as the guest expert at Multi-Sensor Data Fusion seminars that took place at various locations around the country in late 2006 and early 2007. The seminars focused on the latest advancements in multi-sensor data fusion, including both higher and lower-level fusion solutions.
An upcoming seminar is scheduled for April 12-13, 2007, in Las Vegas, Nevada.
The seminars provide guidelines for using various models and techniques to deal with higher level problems associated with decision making in complex, uncertain environments. Examples and demonstrations are drawn from a broad range of critical operational scenarios – from urban operations, to anti-terrorism, air operations, missile defense, and platform/system health monitoring. Available software tools are discussed, and participants engage in an analyses of several examples of military scenarios, including building appropriate belief networks for assessing enemy situations and developing appropriate Bayesian response recommendations.
|
|
|
|
|
On March 23, 2007, Jennifer Roberts presented her research on Bayesian networks for cardiovascular monitoring at Charles River Analytics. “As a step toward developing next-generation cardiovascular monitoring systems for the Intensive Care Unit, I examine Bayesian Networks as a way of organizing incoming patient data and estimating internal patient state,” Roberts explained. “I present a simple Bayesian Network model of the cardiovascular system and show how the model has been used to estimate invasively-measured patient parameters, given data available in one Intensive Care Unit patient record. Using one patient record and samples from a 180-patient database, I demonstrate that the model is capable of capturing patient dynamics by adapting to a specific patient record, provide anecdotal evidence that naïvely incorporating general population data will not improve results, and identify challenges to this type of approach.”
Jennifer Roberts is a graduate student at MIT, who received her Masters on Bayesian Networks for Cardiovascular Monitoring in 2006. Last summer, she worked at Charles River Analytics, and she is currently beginning her PhD with Professor Patrick Winston. Her research interests include structured knowledge representations and concept learning.
|
|
|
|
|
|
BNet News is published by Charles River Analytics, the company that brings you BNet.Builder and BNet.EngineKit.
|
|
|
|
|
|
