This process repeats for each resulting segment until it reaches a "leaf" or terminal node that represents a final prediction or grouping.
The CART algorithm follows a "divide and conquer" strategy to build a model that predicts a target variable based on various input predictors.
It sounds like you're asking about (Classification and Regression Trees) within Minitab , and possibly its integration with solid features (likely referring to solid data structures, stable model features, or solid decision rules).
It automatically finds the optimal number of terminal nodes to maximize predictive ability while avoiding overfitting. Classification vs. Regression Trees
