Abstract Scope |
Material science research is crucial to understand and apply the properties of matter to be able to connect the structure of a material with its performance in application.
However, analyzing datasets to reach such conclusions has proven to be a strenuous, difficult process. For this project I used a dataset for a Caesium-Nitrogen soldering alloy with 25000 indentations performed, and compared the clustering algorithmic compatibility for the nanoidentation mapping data. The clustering plots for K-Means were more precise and meticulous than the other clustering methods studied, proving to be a recommended procedure. To aid the process of clustering for material science and data science users, and to understand various clustering methods as they apply to material science properties, I am helping build an interface that incorporates interactive data visualization. Identifying highly efficient clustering methods and by making this interface accessible to the general public can expedite research for material scientists. |