Abstract Scope |
Using past experimental data is advantageous in material design; however, constructing machine learning models based on this data remains a challenge for materials scientists unfamiliar with machine learning. While data scientists can build appropriate machine learning models with their expertise in statistics, machine learning, and computer science, they may need more domain knowledge, particularly tacit knowledge in materials design. Therefore, collaboration between data scientists and materials scientists is necessary, but the differences in expertise between the two groups can hinder the development of materials informatics. In our corporation, we have leveraged the remote work opportunities caused by Covid-19 to accelerate the usage of materials informatics. We accomplished this by implementing web applications with a user-friendly Graphical User Interface and distributing statistical data analysis tools for materials scientists. As a result, remote work has allowed materials scientists to focus on data organization, statistical data analysis, and the usage of web applications. |