Introduction: Distribution information

From DSL

Hardware and software requirements

Disk space

Full installation of GeNIe may require as much as 5 MB of disk space. The program itself is less than 1 MB large, but the on-line help is quite extensive and includes many graphs, resulting in almost 2 MB large file. The remainder depends on your current computer configuration and availability of run-time libraries.

Memory

GeNIe has practically no minimum memory requirements and can run under a minimum Windows configuration. The actual memory requirement will depend on the size and complexity of the models that you create. Too little memory may result in decreased performance. In general, conditional probability tables grow exponentially with the number of parents of a node. The maximum number of parents of a node will, therefore, determine memory requirements. In addition, memory requirements of the clustering algorithm grow with the connectivity of the network.

Operating system

GeNIe runs under Windows family operating systems (98/NT/2000/XP). The on-line help system requires Microsoft Internet Explorer 4.0 or later. Installation of GeNIe under Windows NT, 2000 and XP operating systems may require administrator privileges.

GeNIe version

To determine the version of GeNIe that you have installed, select About GeNIe from the Help menu. The version number is listed in the small frame of the following window:

Files

GeNIe 2 Complete Setup is an executable file that will install the program, help files and example networks. The setup creates the following files in the destination directory designated by you in the installation process. In addition to creating these files, it also adds GeNIe to the Start Menu under the program group GeNIe 2.0

genie.exe

GeNIe executable file. To start GeNIe, double-click on this file or drag and drop a GeNIe network into the icon of this file.

geniehelp.chm

GeNIe on-line help file in compiled HTML format. You can start on-line help from inside GeNIe (see Help menu) or start it outside of GeNIe by double-clicking on it.

readme.txt

A short text file containing basic information useful in installing or running GeNIe or latest news not included in the on-line help.

tips.txt

GeNIe tip-of-the-day file. Each line in this file contains a simple tip displayed by GeNIe at the start of the program or when invoked from the Help menu.

Uninst.isu

Information needed by the operating system to uninstall GeNIe from your computer.

Example Networks

GeNIe installation creates a directory named Example Networks, containing the following models:

A.xdsl

The A network is a randomly generated network contributed to the community by Alex Kozlov and used for several benchmarks in the UAI literature. Even though A contains only 54 nodes, it is rather densely connected and is known to be daunting to exact algorithms. The network has been first mentioned in (Kozlov & Singh 1996).

Alarm.xdsl

The Alarm network has been developed for on-line monitoring of patients in intensive care units and generously contributed to the community by Ingo Beinlich and his collaborators. The model has first appeared in (Beinlich et al. 1989).

Animals.xdsl

A simple animal guessing game modeled by and made available to the community by Noetic, Inc., the developers of Ergo. The network will guess which of the five animals you have in mind, as you provide information about habitat and characteristics of the animal. The network illustrates well the interaction between probability and propositional logic.

Asia.xdsl

This is an example graphical model useful in demonstrating basics concepts of Bayesian networks in diagnosis. It first appeared in (Lauritzen & Spiegelhalter 1988).

AsiaSmoking.xdsl

This is the Asia network due to Lauritzen and Spiegelhalter (1988), extended to an influence diagram. It is useful in demonstrating basics concepts of influence diagrams. The diagnostic part (the decision and the value node) helps in deciding whether or not to smoke.

B.xdsl

The B network is a randomly generated network contributed to the community by Alex Kozlov and used for several benchmarks in the UAI literature. Even though B contains only 18 nodes, it is rather densely connected and is known to be daunting to exact algorithms. The network has been first mentioned in (Kozlov & Singh 1996).

Coma.xdsl

The Coma or Cancer network, as it is popularly known in the literature, appeared first in Greg Cooper's doctoral dissertation (Cooper 1984).

Credit.xdsl

A simple network for assessing credit worthiness of an individual, developed by Gerardina Hernandez as a class homework at the University of Pittsburgh.

Disease-Test.xdsl

A simple network useful in demonstrating the simplest possible application of Bayes theorem.

Hailfinder2-5.xdsl

Hailfinder is a normative system that forecasts severe summer hail in northeastern Colorado. It has been generously contributed to the community by Ward Edwards and Bruce Abramson. Hailfinder was described in (Abramson et al. 1996). More information about the Hailfinder project can be found at [[1]]

MAUNetwork.xdsl

A simple, incomplete diagram demonstrating Multi-Attribute Utility nodes in GeNIe.

Pinball.xdsl

An influence diagram has been developed by Michael T. Filipiak, Charles Neville and Joseph C. Wynn as a class project at Carnegie Mellon University. It models a real business decision related to a hobby (collecting pinball machines) turned into a business.

Tank.xdsl

A simple network for diagnosing possible explosion in a tank, developed by Gerardina Hernandez as a class homework at the University of Pittsburgh.

tutorial3.xdsl

This is a simple Bayesian network you will create in Tutorial 3: Building Bayesian Networks.

tutorial4.xdsl

This is an example of an influence diagram you will create in Tutorial 4: Creating Influence Diagrams.

tutorial10.xdsl

A tutorial model demonstrating how to control values in GeNIe. It will be created in Tutorial 10 : Controlling Values

tutorial11.xdsl

A tutorial model demonstrating Value Of Information in GeNIe. It will be created in Tutorial 11 : Value Of Information

tutorial12.xdsl

A tutorial model demonstrating sensitivity analysis in GeNIe. It will be created in Tutorial 12 : Sensitivity Analysis

Umbrella.xdsl

A run of the mill umbrella problem, found in most decision analysis textbooks with a weather forecast.

VentureEVPISub.xdsl

A tutorial network demonstrating the use of submodels in GeNIe.