热源布局数据集 (HSLD)

热源布局数据集(HSLD)是一个大规模、多样化的数据集，专门为热源布局预测任务(HSL-TFP)设计。该数据集通过两种随机布局采样方法和几种特殊的布局采样策略生成，确保了样本的高多样性和高质量。HSLD为研究社区提供了一个基准资源，用于开发和评估最先进的深度代理方法，以及基于现有深度回归模型的代理模型的基准结果。数据集的构建考虑了工程适用性，包括样本生成、数据集评估、深度神经网络模型和代理性能指标的全面研究。

The Heat Source Layout Dataset (HSLD) is a large-scale and diverse dataset specifically designed for the Heat Source Layout Prediction Task (HSL-TFP). It is generated via two random layout sampling methods and several special layout sampling strategies, ensuring high diversity and quality of the samples. HSLD provides a benchmark resource for the research community to develop and evaluate state-of-the-art deep surrogate methods, as well as to obtain benchmark results of surrogate models based on existing deep regression models. The construction of the dataset takes into account engineering applicability, including comprehensive studies on sample generation, dataset evaluation, deep neural network models, and surrogate performance metrics.