Open gplobo opened 4 years ago
This histograms shows the probability distribution of lead being present in drinking water. The upper figure shows the measured data, while the lower one shows the modeled data using random forest. The model was trained with a 40% of the total data and provided excellent results (see confusion matrix in other asset). The histogram categories represent the following: 0: P(lead)=0. 1: P(lead)<0.2. 2: P(lead)<0.4. 3: P(lead)<0.6. 4: P(lead)<0.8. 5: P(lead)=1.