Difference between revisions of "Graph Structure of Neural Networks"
From statwiki
(→Related Work) |
|||
| Line 5: | Line 5: | ||
= Introduction = | = Introduction = | ||
| − | = | + | = Major Conclusions = |
| + | |||
| + | (1) graph structure of neural networks matters; | ||
| + | |||
| + | (2) a “sweet spot” of relational graphs lead to neural networks with significantly improved predictive performance; | ||
| + | |||
| + | (3) neural network’s performance is approximately a smooth function of the clustering coefficient and average path length of its relational graph; | ||
| + | |||
| + | (4) our findings are consistent across many different tasks and datasets; | ||
| + | |||
| + | (5) top architectures can be identified efficiently; | ||
| + | |||
| + | (6) well-performing neural networks have graph structure surprisingly similar to those of real biological neural networks. | ||
= Section 1 = | = Section 1 = | ||
Revision as of 15:35, 10 November 2020
Contents
Presented By
Xiaolan Xu, Robin Wen, Yue Weng, Beizhen Chang
Introduction
Major Conclusions
(1) graph structure of neural networks matters;
(2) a “sweet spot” of relational graphs lead to neural networks with significantly improved predictive performance;
(3) neural network’s performance is approximately a smooth function of the clustering coefficient and average path length of its relational graph;
(4) our findings are consistent across many different tasks and datasets;
(5) top architectures can be identified efficiently;
(6) well-performing neural networks have graph structure surprisingly similar to those of real biological neural networks.
