Algorithm 1: Botnet detection using one-class KNN
Input: datasets d1, d2, d3
1.	Convert datasets d1, d2, d3 into PCAP format
2.	Apply filtering based on source and destination IP
3.	$\mathrm{\text{Perform features extraction for}}f\in \left(f_1,f_2,f_3,\dots f_n\right)$
4.	Data preprocessing to eliminate missing, infinite, NAN and HEX values
5.	Perform feature selection using
6.	Filter method
7.	Wrapper method
8.	End feature selection
9.	For each dataset d1, d2, d3 apply one-class KNN
10.	Load the training dataset
11.	Choose the value of k
12.	Train the model
13.	Load the test dataset
14.	For each point in the test data until point = NULL
15.	Find Euclidian distance d to all training data points $d=\sqrt{{\displaystyle \sum }_{i=1}^k{\left(x_i-y_i\right)}^2}$
16.	Store d in a list L and sort it
17.	Choose the first k points
18.	Assign class to the test points
19.	End For
20.	End For