chgen 0.0.1

Comparison of populations using convex hull analysis

Convex Hull Generalization Package

This package provides an approach for comparison of two datasets using convex hull analysis. An intersection value is computed as the amount of points of dataset A that are covered by the convex hull of dataset B in relation to the whole amount of points in dataset A. High intersection values may be an indication of the similarity between the sets. Low intersection values demonstrate that there are features with which the datasets can be distinguished. Additionally, the features found to be important for discriminating between the sets can be compared to those found using standard machine learning classifiers.

More to the package and example of usage at https://git.rwth-aachen.de/jrc-combine/chgen/