Network Pseudohealth Information Recognition Model: An Integrated Architecture of Latent Dirichlet Allocation and Data Block Update
Algorithm 2
Semantic deviation detection algorithm.
Input: : instance information data flow, : information semantic offset detector, : number of integrated members, : instance buffer with size ,: classification quality measurement, : number of instances;
Output: : offset detector integration with 1 classifier and -class weighted classification;
(1)
For all instance do
(2)
Gradually replace with
(3)
(4)
if or the offset is detected, then;
(5)
According to W and , candidate classifier is constructed and empowered;
(6)
According to and , the classifier in integration is empowered;
(7)
if , then ;
(8)
Else if , then replace the weakest block in the integration with ;
