This project implements the Evolutionary State Graph Recurrent Networks proposed in [1], which is a GNN-based method for time series modeling.
Code is compatible with tensorflow version 1.1.0 and Pyhton 3.6.2.
Some Python module dependencies are listed in requirements.txt, which can be easily installed with pip:
pip install -r requirements.txt
An example data format is given where data is stored as a list containing 4 dimensionals tensors such as [number of samples × segment number × segment length × dimensionality].
python run.py -h
usage: run.py [-h] [-v STATENUM] [-d {earthquake,webtraffic}]
[-lr LEARNING_RATE] [-b BATCHSIZE] [-g GPU] [-p MODELPATH]
optional arguments:
-h, --help show this help message and exit
-v STATENUM, --statenum STATENUM
state number
-d {earthquake,webtraffic}, --dataset {earthquake,webtraffic}
select the dataset
-lr LEARNING_RATE, --learning_rate LEARNING_RATE
learning rate
-b BATCHSIZE, --batchsize BATCHSIZE
batch size
-g GPU, --gpu GPU state number
-p MODELPATH, --modelpath MODELPATH
the path of storing model
[1] Wenjie Hu, Yang Yang, Zilong You, Zongtao Liu, Xiang Ren, 2019, Modeling Combinatorial Evolution in Time Series Prediction, In arXiv:1905.05006v2, 2019
@article{hu2019evolutionarygraph,
author = {Wenjie Hu, Yang Yang, Zilong You, Zongtao Liu and Xiang Ren},
title = {Modeling Combinatorial Evolution in Time Series Prediction},
journal = {CoRR},
volume = {abs/1905.05006},
year = {2019},
url = {https://arxiv.org/abs/1905.05006v2},
archivePrefix = {arXiv},
eprint = {1905.05006},
}
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent