ResNet-50 for 12-Lead Electrocardiogram Automated Diagnosis

Algorithm 2

Amplitude scaling.

	Input: Training data pre-process
	Output: Training data pre-process augmented
(1)	α = 0.1
(2)	X = ECG signal
(3)	ScalingFactor = random.normal (loc = 1.0, scale = α, size = (1, X.shape [1]))
(4)	Noise = matmul (ones ((X.shape [0], 1)), scalingFactor)
(5)	return X  Noise