- Traffic flow prediction based on spatio-temporal fusion
- Dataset characteristics
- Source : Traffic data of Bao 'an District, Shenzhen City from 2019 to 2023 (recorded at midnight every day).
- Dimension : 9-dimensional features (number of light/large vehicles, average speed, headway, etc.).
- Preprocessing : Slide the window to construct the time series (window size =9), and divide the training set (90%) and the test set (10%).