Skip to main content

Advertisement

Table 2 Confusion matrix of results where two outlier classes, clustered and unclustered (see Fig. 2) are defined in the control and matched against results from the four predictors – in green the correct number of points detected outliers/inliers (true positives and negatives respectively): percentages represent the true positive rate

From: Implementation and assessment of two density-based outlier detection methods over large spatial point clouds