issai / SCI_prediction_models

Model Description

This repository contains trained models for 9 sustainability systems.

Each system has its own dedicated folder, named according to the specific sustainability system name. Within each folder, you will find two model checkpoints:

• best.pth — the model with the best validation performance
• last.pth — the final checkpoint saved at the end of training

These checkpoints can be used for further fine-tuning, evaluation, or inference.

Intended Use

The models are intended for research and practical applications related to sustainability system analysis and decision support . They may serve as baselines or pretrained backbones for downstream machine learning tasks within this domain.

Citation

If you use these models or the related dataset in your research, please cite:

Bissarinova, U., Tleuken, A., Varol, H. A., & Karaca, F. (2025). AI for City Districts: Estimating City Districts' Sustainability using Satellite Data [Dataset]. Institute of Smart Systems and Artificial Intelligence & Department of Civil and Environmental Engineering, Nazarbayev University. Hugging Face. https://doi.org/10.57967/hf/6652

BibTeX:

@dataset{bissarinova2025ai_city_districts,
  author       = {Bissarinova, Ulzhan and Tleuken, Aidana and Varol, Huseyin Atakan and Karaca, Ferhat},
  title        = {AI for City Districts: Estimating City Districts' Sustainability using Satellite Data},
  year         = {2025},
  institution  = {Institute of Smart Systems and Artificial Intelligence & Department of Civil and Environmental Engineering, Nazarbayev University},
  publisher    = {Hugging Face},
  doi          = {10.57967/hf/6652},
  url          = {https://doi.org/10.57967/hf/6652}
}