Table 5. MCRMSE results of the sequence models based on the numerical and categorical features and codon encoding method without augmentation.
Model name | Training data | Validation data |
---|---|---|
GRU | 0.158 | 0.201 |
LSTM | 0.133 | 0.206 |
HYBRID | 0.131 | 0.138 |
Weighted average | 0.141 | 0.182 |