Skip to main content
. 2025 Aug 25;17:e66798. doi: 10.2196/66798

Table 2. Model performance comparison results.

Model R 2 MAPEa RMSEb
Double exponential smoothing 0.9994 0.0015 0.0748
ARIMAc 0.9992 0.0025 0.0864
MLPd 0.9884 0.0094 0.3396
LSTMe 0.9830 0.0116 0.4110
CNNf 0.9371 0.0237 0.7903
a

MAPE: mean absolute percentage error.

b

RMSE: root mean squared error.

c

ARIMA: autoregressive integrated moving average.

d

MLP: multilayer perceptron.

e

LSTM: long short-term memory.

f

CNN: convolutional neural network.