Browse Teach Curate

Probabilistic Generative Node Classification in Directed Attributed Graphs

A probabilistic model of a directed graph whose nodes carry attributes and labels, defining degree, label, and neighbor-label distributions so the whole graph has an explicit likelihood. Used as a generative classifier, it predicts an unseen node's label by maximum likelihood or maximum a posteriori estimation from its first-order neighborhood under conditional independence, yielding interpretable, inductive predictions that extend Naive Bayes to graph-structured data.

Soundwave Text

A Probabilistic Model for Node Classification in Directed Graphs Diego Huerta and Gerardo Arizmendi

In this work, we present a probabilistic model for directed graphs where nodes have attributes and labels. This model serves as a generative classifier capable of predicting the labels of unseen node…