BriefGPT.xyz
May, 2023
NervePool: 一个单纯性汇聚层
NervePool: A Simplicial Pooling Layer
HTML
PDF
Sarah McGuire, Elizabeth Munch, Matthew Hirn
TL;DR
该研究提出了一个名为NervePool的基于simplicial complexes的池化层，可用于处理深度学习中的graph-structured data，通过对顶点的聚类分配以及高维运算拓展到了更高维度的simplices。
Abstract
For
deep learning
problems on
graph-structured data
, pooling layers are important for down sampling, reducing computational cost, and to minimize overfitting. We define a pooling layer,
→