As a result, the optimal 263-feature set were selected, achieving a correct prediction rate of 83.27% when evaluated by Jackknife cross-validation test.
The capabilities of the model to make predictions were evaluated in terms of the achieved percentage of correct predictions and by testing the results against the values expected from a model making random predictions.
Since the mid-1990s, maternal serum testing and, later, measurement of fetal nuchal translucency, were successful screening tests, and antenatal screening has achieved higher rates of correct predictions and higher coverage year on year.
Even if the achieved results in terms of correct predictions were good, it has to be highlighted that the choice of such a fixed temporal window might have limited the prediction performance.
The median human predictor made 21 correct predictions and achieved an AAE of 0.732 log S units, both better than any individual.
Thus although the WTM achieves a higher percentage of correct predictions, the MIM makes over twice as many predictions, without a comparable loss of correctness.
If a meta-predictor could combine the correct predictions from each PPP, it would achieve a much higher performance.
In the results, we predict DTIs (targets here are proteins from several groups), and show that our method is capable of correctly predicting individual DTI in 27 53 % of cases (depending on the dataset and the target protein group) using only the single top-ranked prediction for a drug, achieving on average the correct prediction in 46.17 % of cases across the four gold standard datasets.
Participant 7 achieved an RMSE of 1.187 log S units, an R2 of 0.637, and 17 correct predictions, while also ranking the compounds best with a ρ value of 0.867.
