How to use ensemble of models in FM or FFM?
-
16-10-2019 - |
Question
I am using Factorization Machines ( libfm) and also the Field Aware Factorization Machines (libffm) for a kaggle competition. I am currently using the single models of each respectively for prediction. I came to know that we can use ensemble of models for both FM and FFM. Can someone explain how this is done ?
Solution
Most ensembling methods are independent from the internals of individual models, e.g. something like majority voting, bagging or model stacking can be directly applicable to any model.
Note: this was originally a comment to the OP.
Licensed under: CC-BY-SA with attribution
Not affiliated with datascience.stackexchange