zinc250k

该数据集名为锌250k，用于评估集成模型在各种分子性质预测任务上的表现。它报告了包括平均绝对误差（MAE）、均方根误差（RMSE）和决定系数R²在内的多种性能指标。在QED性质预测上，集成模型实现了0.003的MAE，比基准模型ASVAE提高了42%。该数据集包含了250,000个样本，任务是对分子性质进行预测。

This dataset, named Zinc250k, is utilized to assess the performance of ensemble models across diverse molecular property prediction tasks. It reports multiple performance metrics including Mean Absolute Error (MAE), Root Mean Squared Error (RMSE), and the coefficient of determination R². For the QED property prediction task, the ensemble model achieved a MAE of 0.003, outperforming the baseline model ASVAE by 42%. This dataset contains 250,000 samples, with the core task being molecular property prediction.