Presents a distribution-free multilevel vector field anomaly detection technique based on Karhunen-Loeve expansions that forms hypothesis tests without distributional assumptions and detects subtle anomalies missed by PCA in simulations.
Gps-based city- wide traffic congestion forecasting using cnn-rnn and c3d hybrid model
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2022 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Distribution-Free Stochastic Analysis and Robust Multilevel Vector Field Anomaly Detection
Presents a distribution-free multilevel vector field anomaly detection technique based on Karhunen-Loeve expansions that forms hypothesis tests without distributional assumptions and detects subtle anomalies missed by PCA in simulations.