I need to find the best model for an algorithm with this caractheristics...
Description of the algorithm;
1) The logic behind it, is based in the machine learning concept. As much data it ingests as parameter the better results it provides, because it "learns" and build patterns which are reused in similar scenarios.
2) It crosses big volumes of data (similar to a complex nested loop).
Something like :
3) It can get the output of another algorithm via parameter.
4) It can get very big objects with data via parameter (example: a whole product)
5) It can get very big object with data via parameter (example: an array of whole products)
This are the basics till the moment. Can someone recommend me, about an adecuate algorithm model ,please?