ES-CP
A tensor anomaly detection method based on CP decomposition that uses element-wise ℓ₁ sparsity penalties to identify localized anomalous points. It is suited for detecting isolated, scattered deviations in tensor-structured data.
Timeline (last 90 days)
ES-CP employs an element-wise ℓ1 penalty term, with the goal of identifying localized anomalies
Both methods are estimated via alternating optimization, and the sparse components have closed-form solution updates
In the multi-channel forging process case study, ES-CP achieved a macro F1 score of 0.82
CP decomposition retains the structural information across the three dimensions of samples, sensors, and time, enabling simultaneous determination of whether an anomaly has occurred and localization of the sensor and time period where it occurred
All Facts (4)
ES-CP employs an element-wise ℓ1 penalty term, with the goal of identifying localized anomalies
50%UnverifiedBoth methods are estimated via alternating optimization, and the sparse components have closed-form solution updates
50%UnverifiedIn the multi-channel forging process case study, ES-CP achieved a macro F1 score of 0.82
50%UnverifiedCP decomposition retains the structural information across the three dimensions of samples, sensors, and time, enabling simultaneous determination of whether an anomaly has occurred and localization of the sensor and time period where it occurred
50%