The BNMA BN Repository

The Decision Graph is applied for the assessment and optimization of an existing threshold-based debris flow warning system. To model the warning system and compute the technical and inherent reliability, the Bayesian Network, which is the Decision Graph without the utility node, can be applied alone. Paper: <www.era.bgu.tum.de...>.

An extremely simple medical diagnosis network example.

The model examines the role of the landscape, fuel load, fuel moisture and fuel breaks on changing the risk to property in San Diego County, USA.

An SR latch is one of the simplest kinds of electronic memory that can be built with logic gates. See <en.wikipedia.org...> for more information.

Set the evidence for 'R' (Reset) and 'S' (Set) to 'False' for all time steps. Then enter a 'True' in 'S' (Set) for time step 3. The 'output' of NOR1 will be locked to True thereafter. (You can then set 'R' to 'True' in a later timestep to clear it.) Requires GeNIe.

This is a BN learned by CaMML using the UC Irvine auto-mpg data set; citing Bache, K. & Lichman, M. (2013). UCI Machine Learning Repository [<archive.ics.uci.edu...>]. Irvine, CA: University of California, School of Information and Computer Science.

A network that uses meteorological data with expert judgements to forecast weather in Northeastern Colorado.

A system used to model waste water treatment and management

A network modelling symptoms of disease and birth asphyxia in a child.

A preliminary model for the production of beer from Danish malting barley grown without pesticides.

This network represents a probabilistic model for diagnosis based on medical knowledge.

Sample network showing probability of lung cancer given certain factors, alongside some symptoms.

A DBN that exhibits oscillating beliefs over time.