towardsdatascience.com/understanding-graph-convolutional-networks-for-node-classification-a2bfdb7aba7b
1 Users
0 Comments
1 Highlights
1 Notes
Tags
Top Highlights
Similar to data pre-processing for any Neural Networks operation, we need to normalize the features to prevent numerical instabilities and vanishing/exploding gradients in order for the model to converge. In GCNs, we normalize our data by calculating the Degree Matrix (D) and performing dot product operation of the inverse of D with AX
Glasp is a social web highlighter that people can highlight and organize quotes and thoughts from the web, and access other like-minded people’s learning.