Table 1 Summary of the tested 10 models, including the neural network architecture, input data, and the applied optimizations
Model | Model Architecture | Hyper-parameter optimizations | Input data |
|---|---|---|---|
1 | Conv1D | batch normalization, causal padding, reverse weight | COV, SEQ, ONTARGET |
2 | Dense | batch normalization, reverse weight | COV, SEQ, ONTARGET |
3 | Conv1D | causal padding, reverse weight | COV, SEQ, ONTARGET |
4 | Dense | reverse weight | COV, SEQ, ONTARGET |
5 | Conv1D | batch normalization, causal padding | COV, SEQ, ONTARGET |
6 | Conv1D | batch normalization, reverse weight, valid padding | COV, SEQ, ONTARGET |
7 | Conv1D | batch normalization, reverse weight, same padding | COV, SEQ, ONTARGET |
8 | Conv1D | batch normalization, causal padding, reverse weight | COV, SEQ |
9 | Conv1D | batch normalization, causal padding, reverse weight | COV, ONTARGET |
10 | Conv1D | batch normalization, causal padding, reverse weight | SEQ, ONTARGET |