Swin–ConvLSTM LULC Prediction Models and Data

This dataset contains trained deep learning models and supporting data for land-use and land-cover (LULC) prediction using spatiotemporal methods.

The repository includes:
- Swin–ConvLSTM model weights
- Supporting data for Malawi case study

These models support forward temporal validation experiments (2015–2020 → 2024).

For code and implementation details, see the GitHub repository:
github.com