Algorithm 2. Initializing and updating algorithm for uPAC.
Require: m > 0, p > 0, dTh > 0, $\lambda$ > 0
1	Cluster training dataset into m clusters //k-medoids clustering
2	Extract representative particles with percentage p into dataset RP
3	Train LS-SVM predictive model PM with RP
4	while xN is valid do
5	$y_N\leftarrow f(PM,{\mathit{x}}_N)$ //predict label
6	allti←ti + 1 //update unchanging time
7	$I\leftarrow \underset{i}{\arg }\left\{\underset{i}{\min }[\exp (t_i/\lambda )d({\mathit{x}}_N,{\mathit{x}}_i)]\right\}$
8	$D\leftarrow d_{Th}-\exp (t_I/\lambda )d({\mathit{x}}_N,{\mathit{x}}_I)$
9	if D > 0
10	tI←0 //clear unchanging time
11	xI ←xN //update RP
12	Update PM //replcaing with universal incremental LS-SVM
13	end if
14	end while