Skip to main content
. 2024 Nov 5;10(22):e40142. doi: 10.1016/j.heliyon.2024.e40142

Table 1.

Results comparison between the proposed system and the existing models.

Citation Existing Models Study Currency RMSE MAPE
Autoregressive integrated moving average (ARIMA), 302.53 42.239
[4] Long Short-Term Memory (LSTM) BTC-USDT 603.68 87.413
Gated Recurrent Unit (GRU) 381.34 49.808



[48] Least Squares Support Vector Machine (LSSVM) BTC-USD 272.155 37.720
Neural Networks with Backpropagation 540.087 74.935



Proposed ADE-TFT(h=2) 209.74 31.0104
model ADE-TFT(h=4) BTC-USD 178.68 29.3295
ADE-TFT(h=8) 167.12 23.1734