Optimal Feature Selection Tool with Machine Learning(LEW-20350-1)

The design tool was developed to choose the optimum set of critical features of data that should be controlled to achieve desired target variable for a given acceptable uncertainty. The tool can conduct data science approaches such as exploratory data analysis, low variance threshold, filter, wrapper, and embedded feature selection technique. The integrated techniques can be applied to determine the critical features that should be controlled to achieve target properties for material systems.