Table 6 Performance of XGBoost across DR strategies.

From: Lightweight machine learning framework using temporal features for electric vehicle demand response forecasting on edge devices

Model

Strategy

MAE

MSE

MAPE

CV

RMSE

R2

XGBoost

Peak

clipping

3.895088

28.439019

9.24

0.053

0.886076

XGBoost

Valley

filling

3.934991

27.642842

8.83

0.152

0.943762

XGBoost

Load

shifting

4.024712

29.306012

5.71

0.15

0.835013

XGBoost

Strategic

conservation

3.823598

26.583385

13.7

0.105

0.975482

XGBoost

Strategic

load growth

3.769721

27.075451

6.18

0.081

1

XGBoost

Flexible

load shape

4.006232

30.413898

9.15

0.087

0.827308