Can I read Statistical Insights on Spectral Embedding on EtoBox?

Statistical Insights on Spectral Embedding by jiaqi bao is a document available to read on EtoBox.

What is Statistical Insights on Spectral Embedding about?

This paper introduces a generalization of the random dot product graph for spectral embedding, allowing for the interpretation of vector representations of graph nodes as latent position estimates, particularly in cases of heterophilic connectivity and negative eigenvalues. The authors demonstrate that spectral embedding yields consistent latent position estimates with Gaussian error, and they recommend using Gaussian mixture models for spectral clustering instead of K-means. Empirical results indicate impr

Author

jiaqi bao

Language

EN