Network Pseudohealth Information Recognition Model: An Integrated Architecture of Latent Dirichlet Allocation and Data Block Update
Algorithm 1
Dataset block integration algorithm.
Input: : an instance information dataset divided into blocks d; K: the number of members of the dataset; B: an instance buffer with a size of d; : classification quality measurement.
Output: : the integration of the classifier weighted as .
(1)
Information data block do for all
(2)
According to and , candidate classifier is established and empowered;
(3)
According to and , all classifiers in set are empowered;
(4)
if , then ;
(5)
Else if , then replace the blocks in the weakest set with ;
