Pathfinder 3: A Bayesian Network with 109 Nodes and 195 Arcs

The Pathfinder Bayesian Network sample 3 is a discrete, very large graph structure for expert systems. It contains 109 nodes, 195 arcs, and 72,079 parameters, with an average Markov blanket size of 3.82. The network was authored by D. Heckerman, E. Horwitz, and B. Nathwani, with foundational research published in 1992.

Use Cases

• Benchmarking inference algorithms based on the network's 72,079 parameters.
• Studying graph topology and connectivity based on the 195 arcs and 109 nodes.
• Evaluating expert system performance based on the discrete, very large network structure.
• Analyzing Markov blanket interactions based on the average size of 3.82.

Strengths

• Well-defined graph structure with 109 nodes and 195 arcs.
• Large number of parameters (72,079) for modeling complex probabilistic relationships.
• Clear academic provenance with authors and a cited 1992 publication.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain tasks.
• Last update date is unknown; freshness unverified.

Provenance

Source

bnlearn Bayesian Network Repository, OpenML.

Collection Method

Likely constructed for the Pathfinder expert system project in medical diagnosis.

Time Range

Dataset associated with research published in 1992.

Freshness

Last updated date is unknown.

Geography

Geographic coverage is unknown.