Table 1 Validation matrices of all the generated multifeature ML-QSAR prediction models.
From: Machine learning driven web-based app platform for the discovery of monoamine oxidase B inhibitors
Prediction Model | PubChemFP prediction model (Training) | PubChemFP prediction model (Test) | SubFP prediction model (Training) | SubFP prediction model (Test) | 1D and 2D mol. desc prediction model (Training) | 1D and 2D mol. desc prediction model (Test) |
|---|---|---|---|---|---|---|
Correlation coefficient | 0.9863 | 0.9397 | 0.9796 | 0.9531 | 0.9852 | 0.8803 |
MAE | 0.1645 | 0.2389 | 0.1683 | 0.2104 | 0.1874 | 0.3068 |
RMSE | 0.212 | 0.2971 | 0.2288 | 0.3023 | 0.2452 | 0.372 |
RAE | 16.9107% | 29.3921% | 18.3017% | 25.0749% | 19.691% | 45.1193% |
RRSE | 19.1019% | 33.0025% | 21.6219% | 31.3838% | 23.0722% | 47.8062% |
Total Number of Instances | 168 | 56 | 163 | 51 | 167 | 52 |
