A Few-Shot Malicious Encrypted Traffic Detection Approach Based on Model-Agnostic Meta-Learning

Algorithm 1

MAML for few-shot malicious encrypted traffic detection.

	Require: distribution over tasks
	Input:, Step size hyperparameter
(1)	Randomly initialize the parameter for the basic learner
(2)	while not done do
(3)	Sample batch of tasks ∼
	for alldo
(4)	Compute the loss function on the support set
(5)	Update adapted parameters by gradient descent algorithm
(6)	end for
(7)	Compute the loss value on the validation set for all tasks
(8)	Update initialization parameters
(9)	end while