Table - PMC

Skip to main content

View full-text article in PMC

. 2020 Oct 12;20(20):5766. doi: 10.3390/s20205766

Subroutine 7 Parallel Relation Table Anomaly Detection Algorithm

Input: Model $M = {E, E_{n}, E_{r}, L P, T_{0}, \dots, T_{| L P |}}$ , test sequence $σ$

Output: Detect if $σ$ is normal or not

1:
Extract the event set $\hat{E}$ of $σ$ .
2:
if $\hat{E} ⊉ E_{n}$ or $E ⊉ \hat{E}$ then
3:
return $σ$ is abnormal.
4:
end if
5:
for event pair ( $e_{i}, e_{i + 1}$ ) in $σ$ do
6:
if $T_{0} (e_{i}, e_{i + 1})$ is ∅ then
7:
return $σ$ is abnormal.
8:
end if
9:
end for
10:
for Loop pattern $l p_{i}$ in $L P$ do
11:
$σ (l p_{i})$ = Sequence processing( $σ, l p_{i}$ )
12:
for event pair ( $e_{i}, e_{i + 1}$ ) in $σ (l p_{i})$ do
13:
if $T_{i} (e_{i}, e_{i + 1})$ is ∅ then
14:
return $σ$ is abnormal.
15:
end if
16:
end for
17:
end for
18:
return $σ$ is normal.