stadlmax/Graph-Posterior-Network

Graph Posterior Network: Bayesian Predictive Uncertainty for Node Classification (NeurIPS 2021)

42
/ 100
Emerging

This project helps machine learning practitioners who build classification models on graph-structured data. It takes in existing graph datasets where entities (nodes) are connected and helps you train models that classify these entities while also understanding how certain or uncertain those classifications are. This allows data scientists or researchers to deploy models with a better grasp of potential errors, especially when the input data might be slightly different from what the model was trained on.

No commits in the last 6 months.

Use this if you need to classify nodes within a graph and require a robust estimate of the predictive uncertainty for each classification, especially in scenarios with shifts or perturbations in the graph structure or node features.

Not ideal if your primary goal is simple, high-throughput node classification without needing detailed uncertainty quantification or if you are working with non-graph data.

graph-data-analysis node-classification predictive-uncertainty machine-learning-reliability out-of-distribution-detection
Stale 6m No Package No Dependents
Maintenance 0 / 25
Adoption 8 / 25
Maturity 16 / 25
Community 18 / 25

How are scores calculated?

Stars

45

Forks

14

Language

Python

License

MIT

Last pushed

Oct 26, 2022

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/stadlmax/Graph-Posterior-Network"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.