Data for "Molecular Simulation-derived features for machine learning predictions of metal glass forming ability"

This is a set of data for the paper "Molecular Simulation-derived features for machine learning predictions of metal glass forming ability".<br>The data includes a molecular dynamics simulated database of critical cooling rates for 11 binary metallic alloys of compositions ranging from 0-10% of the minor alloying element. GFA inspired features have been extracted from the cooling runs and are included along with critical cooling rates in Compiled_Rc_V5.csv<br><br>Full machine learning outputs from the MAST-ML code package are also included for reference in the MASTML_Runs.zip file.<br><br>The 3 files beginning with "lasso..." and the last file "summary..." extract a few key results from the MASTML_Runs.zip file which were used for figures in the paper and in the case of the summary file, for easier comparison of the various model types explored but not detailed in the paper.<br>本数据集配套于论文《基于分子模拟特征的金属玻璃形成能力机器学习预测》。 数据集包含11种二元金属合金的临界冷却速率分子动力学模拟数据库，这些合金的次要合金元素占比范围为0至10%。从冷却模拟流程中提取得到的受玻璃形成能力（Glass Forming Ability，GFA）启发的特征，已与临界冷却速率一同收录于Compiled_Rc_V5.csv文件中。 MAST-ML代码包生成的完整机器学习输出结果也已收录于MASTML_Runs.zip压缩文件中，以供参考。 3个以“lasso”开头的文件以及最后一个名为“summary”的文件，从MASTML_Runs.zip中提取了部分关键结果，这些结果用于论文中的配图；其中summary文件还可用于更便捷地对比论文中探索但未详细阐述的各类模型类型。