Adaptive Decision Spatio-Temporal Neural ODE for Traffic Flow Forecasting with Multi-Kernel Temporal Dynamic Dilation Convolution
PEMSD3: http://pems.dot.ca.gov/
PEMSD4: http://pems.dot.ca.gov/
PEMSD7: http://pems.dot.ca.gov/
PEMSD7L: http://pems.dot.ca.gov/
PEMSD7M: http://pems.dot.ca.gov/
PEMSD8: http://pems.dot.ca.gov/
METR-LA: https://paperswithcode.com/dataset/metr-la
PeMS-BAY: https://paperswithcode.com/dataset/pems-bay
PEMSD(4,8)(3,5,7): change from utils ---> from random_access
Some of these model comparison experiments were conducted on the LibCity(https://bigscity-libcity-docs.readthedocs.io/en/latest/index.html) framework.
run the run_ode.py file.
- python 3.7
- torch 1.7.0+cu101
- torchdiffeq 0.2.2
- fastdtw 0.3.4
- pip install fastdtw==0.3.4