Network Engineering Stack Exchange is a question and answer site for network engineers. It only takes a minute to sign up.
Sign up to join this community
SO on page 6 of this paper it says:
Furthermore, since there is the possibility that the NN attempts to learn classifying the packets using their IP addresses, as the dataset is captured using a limited number of hosts and servers, we decided to prevent this over-fitting by masking the IP addresses in the IP header.
What are the processes to mask the IP exactly? Does it put zeros? or remove it?
What are the processes to mask the IP exactly? Does it put zeros? or remove it?
The idea of "masking" is to hide that piece of data from the program that's analyzing the dataset. The exact method isn't stated. You'd have to ask the authors.
