This will require 2 pair of querysets for known good and known bad instances. The utility function will train based on a sample from one pair, and test accuracy against the second pair. The function will loop for X iterations, or until 100% accuracy (or near 100% accuracy based on a setting/argument).
This will require 2 pair of querysets for known good and known bad instances. The utility function will train based on a sample from one pair, and test accuracy against the second pair. The function will loop for X iterations, or until 100% accuracy (or near 100% accuracy based on a setting/argument).