Is there a way to implement it in

mlforecast

? Thank you!

Hi @Grady Matthias Oktavian, You can forecast several series with a LGBM model, in fact that is one of the reasons machine learning and neural network forecasting has been doing really well in the forecasting competitions. Their ability to leverage shared information across time series with a single global model. jmoralez made an M5 example of the

mlforecast

library, you may want to create a discussion or issue to reach him.

This article might help you: https://medium.com/towards-data-science/forecasting-with-machine-learning-models-95a6b6579090