Towards Spatio-Temporal Aware Traffic Time Series Forecasting

This is a PyTorch implementation of ST-WA in the following paper:
Razvan-Gabriel Cirstea, Tung Kieu, Chenjuan Guo, Shirui Pan, Bin Yang. Towards Spatio-Temporal Aware Traffic Time Series Forecasting.

Requirements

• torch
• scipy>=0.19.0
• numpy>=1.12.1
• pandas>=0.19.2
• pyyaml
• statsmodels
• torch
• tables
• future

Dependency can be installed using the following command:

pip install -r requirements.txt

Data Preparation

The traffic data files are vailable here.

Run the Model

To train the model on different datasets just use the command:

python train.py 

By default it will run the experiments on PEMS4 dataset. To select another dataset open run.py and modify DATASET = 'PEMSX' where X is one of the datasets [3,4,7,8].

The configurations file are located in the config directory. For changing any of the hyper-parameters modify the conf file associated with the dataset and rerun the above command.