What is the CV model of LightGBM (lgb.cv) and how exactly do I use it?

Question

I am able to train a lgmb model using lgb.train and I can do the same with the CV model.

However, I can atleast use the train model for predictions, I am not sure how to understand what the lgb.cv returns.

Answer 1

Check the official documentation here

Specifically, the returned value is the following:

Returns:

eval_hist – Evaluation history. The dictionary has the following format: {‘metric1-mean’: [values], ‘metric1-stdv’: [values], ‘metric2-mean’: [values], ‘metric2-stdv’: [values], …}.

Return type:

dict

A very similar topic is discussed here: Cross-validation in LightGBM