Spatial Shortcuts in Graph Neural Controlled Differential Equations
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866929558566993920 |
|---|---|
| author | Detzel, Michael Nobis, Gabriel Ma, Jackie Samek, Wojciech |
| author_facet | Detzel, Michael Nobis, Gabriel Ma, Jackie Samek, Wojciech |
| contents | We incorporate prior graph topology information into a Neural Controlled Differential Equation (NCDE) to predict the future states of a dynamical system defined on a graph. The informed NCDE infers the future dynamics at the vertices of simulated advection data on graph edges with a known causal graph, observed only at vertices during training. We investigate different positions in the model architecture to inform the NCDE with graph information and identify an outer position between hidden state and control as theoretically and empirically favorable. Our such informed NCDE requires fewer parameters to reach a lower Mean Absolute Error (MAE) compared to previous methods that do not incorporate additional graph topology information. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_19673 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Spatial Shortcuts in Graph Neural Controlled Differential Equations Detzel, Michael Nobis, Gabriel Ma, Jackie Samek, Wojciech Machine Learning I.2.4; G.1.7; G.2.2; G.3; I.6 We incorporate prior graph topology information into a Neural Controlled Differential Equation (NCDE) to predict the future states of a dynamical system defined on a graph. The informed NCDE infers the future dynamics at the vertices of simulated advection data on graph edges with a known causal graph, observed only at vertices during training. We investigate different positions in the model architecture to inform the NCDE with graph information and identify an outer position between hidden state and control as theoretically and empirically favorable. Our such informed NCDE requires fewer parameters to reach a lower Mean Absolute Error (MAE) compared to previous methods that do not incorporate additional graph topology information. |
| title | Spatial Shortcuts in Graph Neural Controlled Differential Equations |
| topic | Machine Learning I.2.4; G.1.7; G.2.2; G.3; I.6 |
| url | https://arxiv.org/abs/2410.19673 |