Skip to main content

Table 1 In each run, a random combination of hyperparameters is tested

From: Load forecasting for energy communities: a novel LSTM-XGBoost hybrid model based on smart meter data

Hyperparameter Value
Past input timesteps as multiple of K 1, 2, 3
Batch size 32, 64
Epochs 50, 75, 100
Steps per epoch 75, 100, 125, 150, 200
Learning rate 0.1, 0.01, 0.005, 0.001
Number of units in LSTM-layer 20, 70, 130, 200, 260
Number of units in dense layers 20, 70, 130, 200, 260