A-NeSI: A Scalable Approximate Method for Probabilistic Neurosymbolic Inference.
Emile van KriekenThiviyan ThanapalasingamJakub M. TomczakFrank van HarmelenAnnette ten TeijePublished in: CoRR (2022)
Keyphrases
- detection method
- significant improvement
- pairwise
- computational cost
- high precision
- prior information
- cost function
- probabilistic model
- feature set
- inference process
- synthetic data
- segmentation algorithm
- support vector machine svm
- generative model
- model selection
- mutual information
- similarity measure
- high accuracy
- knn
- prior knowledge
- multiscale