Create some kind of visualizer that enables the user to steer the outlier removal/pruning process through visualizations.
For a single dimension, this would depict an iterative outlier pruning process, for instance using z-scores and eliminating anything outside of 3z. For multidimensional data, perhaps select the feature with the most variance. Show how the variance of the target is changing throughout the outlier removal process.
Create some kind of visualizer that enables the user to steer the outlier removal/pruning process through visualizations.
For a single dimension, this would depict an iterative outlier pruning process, for instance using z-scores and eliminating anything outside of 3z. For multidimensional data, perhaps select the feature with the most variance. Show how the variance of the target is changing throughout the outlier removal process.