PriorGuide enables efficient incorporation of arbitrary priors at inference time for amortized diffusion-based simulation-based inference, without retraining the model.
We propose an architecture extension to existing tabular foundations to generate joint predictive samples ~20x faster with minimal increase in training overhead and drop in performance.
We propose a formal framework for probabilistic multi-dimensional classification.
We develop Bayesian structure scores for deterministic PCs, i.e., the structure likelihood with parameters marginalized out.
We propose Generalized Bayesian Network Classifier (GBNC), a generalized framework for solving probabilistic multi-dimensional classification problems with complex types of input.