DP-SUM-CUSUM adds calibrated Laplace noise to summed CUSUM statistics to achieve sequential ε-differential privacy for multi-stream change detection while providing explicit bounds on false-alarm run length and worst-case detection delay.
Online privacy- preserving data-driven network anomaly detection
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.ST 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Sequential Change Detection for Multiple Data Streams with Differential Privacy
DP-SUM-CUSUM adds calibrated Laplace noise to summed CUSUM statistics to achieve sequential ε-differential privacy for multi-stream change detection while providing explicit bounds on false-alarm run length and worst-case detection delay.