[Python] Optimal Approach for Training an AI Model to Correct Errors in Multipolygon...

I need to select an AI model and a Python library that would be the most optimal for training. I have coordinates represented as a Multipolygon field from the djangorestframework-gis library, and they have small errors within different ranges—approximately 0.75 to 1 meter. Additionally, I have correct coordinates for the same land plots. The model needs to learn to find the difference between the correct and incorrect data and, based on that, discover an algorithm to determine the error for future correction of the incorrect coordinates.

Continue reading...