Deep Learning for Microstructure Property Relationships - Results (MATLAB data)

The data contains trained networks and interpretation data saved into ".MAT" files. The framework was orginally deployed on MATLAB R2022b. It has also been tested to be functional on MATLAB R2021a. The data used in this work has been collected from various literature which may contain copyrighted materials. Therefore we are unable to provide the data used for training. However, we have provided all the data sources in the file "Data_Sources.xlsx" located inside the "1_Alloy_Data" folder. MATLAB scripts for data preprocessing, optimization, training and interpretation of the deep learning framework are provided in a separate git repository: https://github.com/adityag23/Deep-Learning-Microstructure-Property

Requirements:
MATLAB Deep Learning Toolbox
MATLAB Image Processing Toolbox

本数据集包含已训练的网络和以".MAT"文件格式保存的解释数据。该框架最初部署于MATLAB R2022b版本，并经过测试，确认在MATLAB R2021a版本上亦能正常运行。本研究中所使用的数据系源自众多文献，其中可能包含受版权保护的内容。因此，我们无法提供用于训练的数据。然而，我们已将所有数据来源详细列出，存放于“1_Alloy_Data”文件夹内的“Data_Sources.xlsx”文件中。此外，针对数据预处理、优化、训练及深度学习框架解释的MATLAB脚本已另存于独立的git仓库：https://github.com/adityag23/Deep-Learning-Microstructure-Property。使用本数据集需具备MATLAB Deep Learning Toolbox和MATLAB Image Processing Toolbox。