Metrics for the evaluation of learned causal graphs based on ground truth
Publication date
2024-03
Document type
Conference paper
Author
Rehak, Josephine
Falkenstein, Alexander
Doehner, Frank
Beyerer, Jürgen
Organisational unit
Fraunhofer IOSB/Karlsruher Institut für Technologie (KIT)
Fraunhofer IOSB
Part of the university bibliography
✅
Keyword
Causal graph
Metric
Causal discovery
Ground truth
Bayesian network structure learning
Causal structure learning
Acyclic graph
Ancestral graph
Abstract
The self-guided learning of causal relations may contribute to the general maturity of artificial intelligence in the future. To develop such learning algorithms, powerful metrics are required to track advances.
In contrast to learning algorithms, little has been done in regards to developing suitable metrics. In this work, we evaluate current state of the art metrics by inspecting their discovery properties and their considered graphs. We also introduce a new combination of graph notation and metric, which allows for benchmarking given a variety of learned causal graphs. It also allows the use of maximal ancestral graphs as ground truth.
In contrast to learning algorithms, little has been done in regards to developing suitable metrics. In this work, we evaluate current state of the art metrics by inspecting their discovery properties and their considered graphs. We also introduce a new combination of graph notation and metric, which allows for benchmarking given a variety of learned causal graphs. It also allows the use of maximal ancestral graphs as ground truth.
Version
Not applicable (or unknown)
Access right on openHSU
Open access