Supplementary Material for
On the Evaluation of Unsupervised Outlier Detection: Measures, Datasets, and an Empirical Study
by G. O. Campos, A. Zimek, J. Sander, R. J. G. B. Campello, B. Micenková, E. Schubert, I. Assent and M. E. Houle
Data Mining and Knowledge Discovery 30(4): 891-927, 2016, DOI: 10.1007/s10618-015-0444-8

Parkinson (5% of outliers version#05)

The data set consists of medical data distinguishing healthy people from those suffering from Parkinson's disease. The latter were labeled as outliers.

Download all data set variants used (278.6 kB). You can also access the original data. (parkinsons.data)

Normalized, without duplicates

This version contains 22 attributes, 50 objects, 2 outliers (4.00%)

Download raw algorithm results (207.4 kB) Download raw algorithm evaluation table (5.9 kB)

Best Parameters

The following table contains the best (overall and per-method) results for each method and evaluation measure (when the same score was achieved twice, only the smallest k is given).
The Maximum F1-Measure is complimentary in addition to the measures in the original publication.

Algorithm k P@n Adj. P@n AP Adj. AP Max-F1 Adj. MF1 ROC AUC
KNN 1 0.50000 0.47917 0.55882 0.54044 0.66667 0.65278 0.84375
KNNW 1 0.50000 0.47917 0.56667 0.54861 0.66667 0.65278 0.86458
LOF 3 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
SimplifiedLOF 5 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
LoOP 5 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
LDOF 5 0.50000 0.47917 0.45000 0.42708 0.57143 0.55357 0.95833
LDOF 7 0.50000 0.47917 0.83333 0.82639 0.80000 0.79167 0.98958
ODIN 2 0.22222 0.18981 0.22222 0.18981 0.36364 0.33712 0.92708
ODIN 5 0.50000 0.47917 0.27703 0.24690 0.50000 0.47917 0.65104
ODIN 7 0.50000 0.47917 0.52128 0.50133 0.66667 0.65278 0.55208
ODIN 30 0.50000 0.47917 0.54348 0.52446 0.66667 0.65278 0.78646
FastABOD 3 0.50000 0.47917 0.54348 0.52446 0.66667 0.65278 0.78125
FastABOD 6 0.50000 0.47917 0.57692 0.55929 0.66667 0.65278 0.88542
KDEOS 11 0.50000 0.47917 0.50000 0.47917 0.66667 0.65278 0.96875
KDEOS 13 0.50000 0.47917 0.83333 0.82639 0.80000 0.79167 0.98958
LDF 2 0.50000 0.47917 0.70000 0.68750 0.66667 0.65278 0.96875
INFLO 3 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000
COF 4 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000 1.00000

Plots

Precision at n
Adjusted precision at n
Average precision
Adjusted average precision
Maximum F1 score
Adjusted maximum F1 score
ROC AUC
Diversity
A: KNN, B: KNNW, C: LOF, D: SimplifiedLOF, E: LoOP, F: LDOF
G: ODIN, H: KDEOS, I: COF, J: FastABOD, K: LDF, L: INFLO

Not normalized, without duplicates

This version contains 22 attributes, 50 objects, 2 outliers (4.00%)

Download raw algorithm results (208.8 kB) Download raw algorithm evaluation table (11.4 kB)

Best Parameters

The following table contains the best (overall and per-method) results for each method and evaluation measure (when the same score was achieved twice, only the smallest k is given).
The Maximum F1-Measure is complimentary in addition to the measures in the original publication.

Algorithm k P@n Adj. P@n AP Adj. AP Max-F1 Adj. MF1 ROC AUC
KNN 1 0.00000 -0.04167 0.11667 0.07986 0.25000 0.21875 0.65625
KNN 2 0.00000 -0.04167 0.13448 0.09842 0.28571 0.25595 0.67708
KNN 47 0.00000 -0.04167 0.11111 0.07407 0.20000 0.16667 0.75000
KNNW 1 0.50000 0.47917 0.28846 0.25881 0.50000 0.47917 0.73958
LOF 1 0.00000 -0.04167 0.25000 0.21875 0.40000 0.37500 0.87500
LOF 2 0.50000 0.47917 0.52703 0.50732 0.66667 0.65278 0.63542
LOF 3 0.50000 0.47917 0.53333 0.51389 0.66667 0.65278 0.70833
SimplifiedLOF 1 0.50000 0.47917 0.39286 0.36756 0.50000 0.47917 0.93750
SimplifiedLOF 4 0.50000 0.47917 0.55000 0.53125 0.66667 0.65278 0.81250
LoOP 1 0.50000 0.47917 0.39286 0.36756 0.50000 0.47917 0.93750
LoOP 3 0.50000 0.47917 0.54762 0.52877 0.66667 0.65278 0.80208
LoOP 6 0.50000 0.47917 0.55000 0.53125 0.66667 0.65278 0.81250
LDOF 2 0.50000 0.47917 0.32143 0.29315 0.50000 0.47917 0.86458
LDOF 4 0.50000 0.47917 0.59091 0.57386 0.66667 0.65278 0.90625
ODIN 2 0.25000 0.21875 0.17500 0.14062 0.33333 0.30556 0.87500
ODIN 3 0.33333 0.30556 0.19792 0.16450 0.40000 0.37500 0.71875
FastABOD 3 0.00000 -0.04167 0.13571 0.09970 0.28571 0.25595 0.68750
FastABOD 16 0.00000 -0.04167 0.15203 0.11669 0.33333 0.30556 0.60417
KDEOS 3 0.50000 0.47917 0.35000 0.32292 0.50000 0.47917 0.90625
KDEOS 4 0.50000 0.47917 0.55882 0.54044 0.66667 0.65278 0.84375
KDEOS 6 0.50000 0.47917 0.64286 0.62798 0.66667 0.65278 0.94792
LDF 1 0.50000 0.47917 0.29762 0.26835 0.50000 0.47917 0.79167
LDF 2 0.50000 0.47917 0.53125 0.51172 0.66667 0.65278 0.68750
INFLO 2 0.50000 0.47917 0.60000 0.58333 0.66667 0.65278 0.91667
INFLO 49 0.51020 0.48980 0.52000 0.50000 0.66667 0.65278 0.75000
COF 1 0.50000 0.47917 0.39286 0.36756 0.50000 0.47917 0.93750
COF 4 0.50000 0.47917 0.54348 0.52446 0.66667 0.65278 0.78125
COF 7 0.50000 0.47917 0.55000 0.53125 0.66667 0.65278 0.81250

Plots

Precision at n
Adjusted precision at n
Average precision
Adjusted average precision
Maximum F1 score
Adjusted maximum F1 score
ROC AUC
Diversity
A: KNN, B: KNNW, C: LOF, D: SimplifiedLOF, E: LoOP, F: LDOF
G: ODIN, H: KDEOS, I: COF, J: FastABOD, K: LDF, L: INFLO