ETT-17 (Energy Transition Tasks-17)

ETT-17是由麻省理工学院和苏黎世联邦理工学院共同创建的数据集，包含17个与可再生能源增强相关的数据集，涵盖六个不同的应用领域。数据集旨在通过统一的多任务机器学习模型解决与气候变化和能源转型相关的任务。ETT-17数据集包括分布外验证和测试数据，支持模型在零/少样本学习中的表现评估。该数据集的应用领域包括需求预测、发电预测、交通预测、气候和天气模型分辨率提升、催化剂发现以及气候和能源政策分析，旨在通过机器学习技术推动可再生能源的转型和气候变化问题的解决。

ETT-17 is a dataset co-created by the Massachusetts Institute of Technology and ETH Zurich, which consists of 17 sub-datasets related to renewable energy enhancement and covers six distinct application domains. This dataset aims to address tasks associated with climate change and energy transition via unified multi-task machine learning models. The ETT-17 dataset includes out-of-distribution validation and test data, enabling the evaluation of model performance in zero-shot and few-shot learning scenarios. Its application areas cover demand forecasting, power generation forecasting, traffic forecasting, climate and weather model super-resolution, catalyst discovery, as well as climate and energy policy analysis. The dataset is designed to promote renewable energy transition and tackle climate change issues through machine learning technologies.