Fig. 2.

Time Series Cross-validation: The training set increases sequentially, maintaining the temporal order of the data for predicting one month ahead (testing set). The training set divided further into a training set and validation set using 80:20% for monitoring the models performance during training