Skip to content

Evaluating the performance of STEP with WaveNet and Graph WaveNet architectures on multivariate time series forecasting

Notifications You must be signed in to change notification settings

nataliekoh/GNNs_MultivariateTSForecasting

Repository files navigation

Multivariate Time Series Forecasting with Graph Neural Networks

Natalie Koh, Zachary Laswick, Daiwei Shen

Datasets

Architectures Used

  • STEP
  • Graph WaveNet
  • Simple graph convolutional network with LSTM layer implemented in Keras

Scripts

  • For data pre-processing, see PruneDatasets_SingleSubject.ipynb.
  • To run STEP on the datasets, use scripts in STEP/ModifiedSTEPCode.
  • To run Graph WaveNET, cd into the WaveNet directory and run python train.py --gcn_bool.
  • To run the simple GCN implemented in Keras, use KerasGNNwLSTM.ipynb.

About

Evaluating the performance of STEP with WaveNet and Graph WaveNet architectures on multivariate time series forecasting

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published