Algorithm 1 Pseudocode for the proposed outliers detection approaches.
1	Input: T, alpha, apr
2	//Typicality degrees matrix in nxc dimension, and built by an
3	//unsupervised possibilistic clustering algorithm
4	// alpha, threshold possibility value for outlier testing
5	// apr, number of the approach to be used in outlier detection
6	Output: Outliers
7	//Outliers, vector of n length to store the flags of outliers
8	n <- count of rows of matrix T
9	c <- count ofcolumns of matrix T
10	// If alpha is undefined, use 0.05 as the default value
11	if alpha is null then alpha = 0.05
12	Outliers <- {0} //Assign 0 to all elements of the outliers
13	for k = 1 to n do
14	if apr = 1 then
15	sumT <- 0
16	for i = 1 to c do
17	sumT <- sumT + T[i,k]
18	end
19	avgT <- sumT / c
20	if avgT <= alpha then
21	Outliers[k] <- 1
22	end
23	else
24	if apr = 2 then
25	isOutlier <- True
26	for i = 1 to c do
27	if T[i,k] >= alpha then
28	isOutlier <- False
29	end
30	end
31	if isOutlier = True then
32	Outliers[k] <- 1
33	end
34	end
35	end
36	end
37	return Outliers