Introduction to Bayesian Networks Using Netica

Course Description: Innovative Decisions, Inc. is offering our three-day Bayesian networks course May 15-17, 2012 in Vienna, Virginia (a suburb of Washington, DC). The cost is $1,950, which includes all handouts as well as continental breakfast, lunch and snacks each day. IDI is also offering this course in 2012 on September 18-20 and November 13-15.

Discount of $200 for early registration (deadline April 30). We also offer a discount for multiple persons from the same organization (please contact Kellie for details).

The course focus is on constructing and analyzing Bayesian networks (BNs) in Netica - from simple models for learning the basics to more complex, real-world applications for learning advanced features. Through a hands-on learning process, students will have the opportunity to explore examples and work on their own networks during class exercises.

We are offering an optional fourth day for instruction on using Netica through library elements in the ExtendSim discrete event simulation software. This provides the analyst with several new capabilities, including learning of probability tables and utilities from the results of a simulation, testing a network in a simulation environment that represents the real world, and Bayesian networks control of discrete simulations. Bayesian networks can also be used to model highly complex phenomenon within a discrete simulation that cannot be described mathematically. The cost for this option is an additional $650. Please contact Kellie Gallahan to register, or Dave Brown for technical questions.

Location: 1945 Old Gallows Road, Suite 207, Vienna, VA 22182

Time: 8:30am to 4:30pm

Instructors: 

Joseph A. Tatman

Suzanne M. Mahoney

Dennis M. Buede

Amanda Hepler

Planned topics:

Introduction to basic concepts – health screening clinic example

Essential probability for understanding BNs

Inference, combining evidence, updating probabilities:

   Building a first BN – drug testing example

   A more realistic BN – simple car diagnosis example

   Actual application – medical diagnosis 

   Troubleshooting a system using deterministic functions and continuous variables

Determining the value of evidence

Eliciting structure (nodes and arcs) from experts

Eliciting probabilities (conditional probability tables) from experts

Estimating a probability – Beta updating example

Learning probabilities from data

Learning structure from data

Combining expert knowledge with data

Survey of BN applications and common patterns

Intro to advanced topics

Other possible topics:

Named probability distributions in Netica

Decision making in the BN

Combining deterministic mathematical models with BN models

Basics of the propagation algorithm behind BNs

Intro to integrating Netica with Excel

Intro to building a BN web application using Netica

Using the Netica library in the Extend discrete event simulation package

Influence diagrams

Evidential reasoning