|
|
|
|
Welcome to the Visual Tour of Bayesware
Discoverer 1.0! |
|
|
|
This Visual Tour provides an overview of the
basic features of Bayesware Discoverer. It is neither a crash course or a
detailed examination, but rather a taste of the capabilities of the
program. |
|
|
|
The example used in the tutorial is a publicly
available database of loan applicants to a German bank. Bayesware
Discoverer is supplied with a database of 1000 customers of a German bank
applying for a loan. The database records, for each customer, twenty-one
values, such as Age, Marital Status, Credit History, Employment, Residence
Time, Housing conditions and the outcome of the loan. Use the Side bar or
the buttons at the bottom of this page to watch Bayesware Discoverer in
action on these data. |
|
|
|
This Visual Tour relies on AVI movies. Please be
patient while the movies are downloaded on your computer. This Visual Tour
is designed for Microsoft Explorer 4 or higher and your computer screen
should provide a 1024 x 768 resolution or higher. If you do not have
Microsoft Explorer 4, you can download it from the Microsoft website. Click on white board icon on the right
hand corner of your browser to enjoy the full screen presentation.
Alternatively, use the itemized side bar or the controls at the bottom of
the screen. To go back to the Bayesware Discoverer page, click on the
Bayesware Discoverer logo on the top right corner. |
|
|
|
Enjoy the ride! |
|
|
|
|
Let's suppose you have started Bayesware
Discoverer. After the screen splash, a Network Window will appear, as shown
in the picture below. Load now the database. |
|
|
|
|
If you have connected your Network Window to a
database, a Database window will appear and you will be able to edit and
save the content of the connected database |
|
|
|
|
Once the database has been loaded into the
program, you can start the automated modeling process and ask Bayesware
Discoverer to search for the most probable network of dependencies. |
|
|
|
|
|
|
Bayesware Discoverer can be used to explore
various aspects of the extracted model, such as the conditional probability
distributions, their variance, confidence intervals, the likelihood of the
extracted dependency model, and so on. |
|
|
|
|
|
|
Bayesware Discoverer provides a 3D interface to
explore the quantitative components of a Bayesian network. |
|
|
|
|
The network is a self-contained Bayesian network
based decision support system able to predict, explain, and explore different scenarios using
state-of-the-art propagation algorithms. |
|
 |
Notes