Few-Shot Learning-Based Network Intrusion Detection through an Enhanced Parallelized Triplet Network

Table 2

Label distribution of few-shot datasets based on the UNSW_NB15.

Describe	Type	Train	Test
Normal traffic	Normal	70	10000
Known attack traffic	Reconnaissance	10	1000
	Backdoor	10	573
	DoS	10	1000
	Exploits	10	1000
	Analysis	10	667
	Fuzzers	10	1000
	Generic	10	1000
Unknown attack traffic	Worms	0	44
	Shellcode	0	378
—	Total	140	16,662