BNet.News
In This Issue: September/October 2005 
•   Charles River Analytics' Scientist Discusses Bayesian Belief Networks at the 8th International Conference on Information Fusion
•   BNet.Builder User Tip: Working with States
•   BNet.EngineKit User Tip: Working with States
•   Visit Our Web Pages for White Papers on Belief Networks
•   Publicize Your Belief Network Success Story
Charles River Analytics' Scientist Discusses Bayesian Belief Networks at the 8th International Conference on Information Fusion
Dr. Subrata Das, Chief Scientist at Charles River Analytics, was a presenter at the 8th International Conference on Information Fusion, which took place in Philadelphia from July 25-29, 2005. Dr. Das delivered a two-part tutorial at the conference entitled "An Integrated Approach to Data Fusion and Decision Support." In his tutorial, Dr. Das included a detailed presentation on static/dynamic Bayesian belief networks.

Dr. Das's tutorial was divided into two parts, Situation Assessment and Response Recommendation. In the first part of his tutorial, Dr. Das discussed the BNet Bayesian Network tool as a foundational tool in Situation Assessment, as well as Problem Modeling using BNet. In the second part of his tutorial, Dr. Das discussed Bayesian Belief Networks as a Foundational Technology as well as Bayesian Probability as a Aggregation and Decision Making tool.

Dr. Das and his colleagues, Charles River Analytics’ scientists Josh Introne and Dave Lawless, also presented the following related papers at the conference:

"Factored Particle Filtering for Situation Assessment in Urban Environments"

"Truth Maintenance System with Probabilistic Constraints for Enhanced Level Two Fusion"

To read more about Dr. Das's tutorial, "An Integrated Approach to Data Fusion and Decision Support," click here.

Visit Our Web Pages for White Papers on Belief Networks
Visit Charles River Analytics' Web site to read the following white papers on Belief Networks:

"An Introduction to Belief Networks"

"Getting more out of your data with Belief Networks"

Click here to visit our Belief Network Modeling Web page.

BNet.Builder User Tip: Working with States
BNet.Builder makes it easy to work with the states of a node. Both the
expanded node and the Node Properties window contain tools for adding,
renaming, reordering, and removing states.

To add a new state to a node simply click the "Add a new state" button,
which looks like a plus sign. A new state will then be added to the
node.

To rename a state click on the state name to begin editing it. When
you're finished, either hit Enter or click somewhere else to rename the
state.

To reorder a state drag the grip icon, just to the left of the state
name, up or down. As you drag, you'll see a horizontal line where the
state will be placed. Drop the state in the new position to reorder it.

To remove a state first select it by clicking on it, and then click the
"Remove the selected state button" which looks like the letter X. Note
that you cannot remove a state from a node that has only two states,
because every node must have at least two states.

BNet.EngineKit User Tip: Working with States
The BNet.EngineKit API also makes it easy to work with the states of a
node. The DiscreteNode class provides a getStates() method which
returns a StateList object. The StateList interface extends the List
interface of the Java Collections framework, and makes it easy to add,
rename, reorder, and remove states. See the BNet.EngineKit javadocs for
full details.

To add a new state use one of the add methods provided by the List
interface. For example, to add a new state named "new state", do this:

DiscreteNode node = ...;
StateList states = node.getStates();
states.add("new state");


To rename a state use set method provided by the List interface. For
example, to rename the state at index zero to "new state name", do this:

DiscreteNode node = ...;
StateList states = node.getStates();
states.set(0, "new state name");


To reorder a state use the reorder method provided by the StateList
interface. For example, to reorder the state "true" to index one, do
this:

DiscreteNode node = ...;
StateList states = node.getStates();
states.reorder("true", 1);


To remove a state use one of the remove methods provided by the List
interface. For example, to remove the state "medium", do this:

DiscreteNode node = ...;
StateList states = node.getStates();
states.remove("medium");

Publicize Your Belief Network Success Story
Want to show off your applications of belief networks to others using them in their research? Send us your success stories. We would like to include at least one in each edition of BNet.News.

Send us your success story

BNet News is published by Charles River Analytics, the company that brings you BNet.Builder and BNet.EngineKit.  View our Privacy Policy