Free Access

Table 2.

Classifying GCs from the Virgo and Fornax tabular data sets.

Method TPR FPR FDR AUC ROC
Logistic Regression 0.761±0.007 0.054±0.002 0.204±0.005 0.951±0.001
Support Vector Machine (linear) 0.778±0.007 0.053±0.002 0.196±0.005
Support Vector Machine (radial) 0.858±0.005 0.042±0.002 0.151±0.006
Nearest Neighbour 0.897±0.005 0.040±0.002 0.139±0.007
12 Nearest Neighbours 0.914±0.005 0.030±0.001 0.100±0.004 0.987±0.001
Decision Tree 0.892±0.009 0.036±0.003 0.127±0.010 0.983±0.001
Random Forest 0.909±0.005 0.018±0.001 0.066±0.005 0.992±0.001
AdaBoost 0.910±0.005 0.018±0.001 0.065±0.005 0.992±0.001
CatBoost 0.920±0.006 0.025±0.002 0.089±0.007 0.990±0.001
Neural Network (29-1) 0.756±0.007 0.056±0.002 0.210±0.005 0.948±0.001
Neural Network (29-100-100-1) 0.932±0.009 0.028±0.003 0.097±0.010 0.992±0.001

Notes. The reported results are averages over ten random splits (train, validation and test). Uncertainties are given as standard deviations and the used decision threshold is 0.5. Support vector machines and nearest neighbour only provide a class label and hence, no AUC ROC is reported. For neural networks, the number of neurons per layer is given in brackets.

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.