Machine Learning Math Idea – Could Something Like This Exist. M(x) Goes To model_infinite() when x Goes To infinite. A model() of the infinity

Maybe there exist a model() of the infinity? That’s my idea anyway.

Inspired by this I got the idea of seamless pooling function. A simple way to get something looking infinetly  large. Then you could have similarity pooling.

Maybe convolutions and poolings should not be so rigid in machine learning models. Just move the data around.