I have datasets with a network of lines and features in an oracle geo database. I also have a dataset with points that are placed by users based on this data.
The goal is to give 2 inputs, an asset and the surrounding network, and get a prediction of where points need to be placed.
I have been searching for examples of this kind of machine learning. But can only find image recognition and interpretation examples. The closest I have come is an example of shape file masks used to predict parking lot shapes. There they used a rough mask and a satellite image, there the result was a new polygon shape of the predicted parking lot. But that is not what I'm looking for.
The solution I see at this point is creating a raster image of the network and use that as an input, but I am sure that that is a round about way of doing it and that would mean the resulting model needs an image as an input as well.
Thank you in advance for your guidance.