pIC50 prediction on pIC50 Generalization - Ion Channel
1.24RMSERandom Forest
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Random ForestArchitecture=RF, Ensemble Strategy=single, Representation=novel representations learned from seq2seq2018.06 | 1.24 | 0.3 | |
| Unified RNN-CNNArchitecture=Unified RNN-CNN, Ensemble Strategy=parameter+NN ensemble, Representation=novel representations learned from seq2seq2018.06 | 1.3 | 0.18 | |
| Separate RNN-CNNArchitecture=Separate RNN-CNN, Ensemble Strategy=parameter+NN ensemble, Representation=novel representations learned from seq2seq2018.06 | 1.33 | 0.18 | |
| Unified RNN-CNNArchitecture=Unified RNN-CNN, Ensemble Strategy=parameter ensemble, Representation=novel representations learned from seq2seq2018.06 | 1.33 | 0.18 | |
| Unified RNN-CNNArchitecture=Unified RNN-CNN, Ensemble Strategy=single, Representation=novel representations learned from seq2seq2018.06 | 1.34 | 0.17 | |
| Separate RNN-CNNArchitecture=Separate RNN-CNN, Ensemble Strategy=parameter ensemble, Representation=novel representations learned from seq2seq2018.06 | 1.36 | 0.18 |