|
Model | MSE | RMSE |
|
Proposed model | 0.000609 | 0.026026 |
Random search: convolutional neural networks [46] | 314.9 | 17.7 |
Particle swarm optimization: convolutional neural networks [46] | 194.5 | 13.9 |
Firefly algorithm: convolutional neural networks [46] | 179.9 | 13.4 |
Modified firefly algorithm: multichannel convolutional neural network [46] | 120.9 | 10.9 |
Genetic algorithm: long short-term memory [47] | 0.007 | 0.089 |
Differential evolution: functional link artificial neural network [47] | 0.009 | 0.03 |
Genetic algorithm: radial basis function [47] | 0.003 | 0.056 |
Cat swarm optimization: auto regressive moving average [47] | 0.001 | 0.038 |
Particle swarm optimization: ELMAN [47] | 0.004 | 0.07 |
Particle swarm optimization: multilayer perceptron [47] | 0.002 | 0.052 |
Biogeography-based optimization: multilayer perceptron [47] | 0.001 | 0.043 |
Cat swarm optimization: multilayer perceptron [47] | 0.0008 | 0.029 |
Gated recurrent unit based on complete ensemble empirical mode decomposition of adaptive noise: wavelet [48] | 496.07 | 22.27 |
Long short-term memory [48] | 693.79 | 26.34 |
Autoregressive integrated moving average [48] | 769.68 | 27.74 |
Gated recurrent unit [48] | 552.99 | 23.51 |
Convolutional neural network: bidirectional long short-term memory [48] | 589.52 | 24.28 |
Support vector machine [49] | 0.003 | 0.058 |
Back propagation neural network [49] | 0.017 | 0.132 |
Traditional convolutional neural networks [50] | 0.68 | 0.824666 |
Denoised traditional convolutional neural networks [50] | 0.468 | 0.683787 |
Traditional convolutional neural networks LightGBM [50] | 0.382 | 0.617769 |
Denoised traditional convolutional neural networks LightGBM [50] | 0.215 | 0.463204 |
ResNet [50] | 0.508 | 0.713024 |
Denoised ResNet [50] | 0.424 | 0.65097 |
ResNet LightGBM [50] | 0.365 | 0.603879 |
Denoised ResNet LightGBM [50] | 0.156 | 0.395185 |
|