NN anomaly detection methods achieve competitive empirical performance on benchmarks and receive finite-sample misclassification guarantees derived from empirical DTM analysis under Huber's contamination model with geometric assumptions.
Efficient algorithms for mining outliers from large data sets
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Statistical Analysis of Nearest Neighbor Methods for Anomaly Detection
NN anomaly detection methods achieve competitive empirical performance on benchmarks and receive finite-sample misclassification guarantees derived from empirical DTM analysis under Huber's contamination model with geometric assumptions.