Pathological Myopia Image Recognition Strategy Based on Data Augmentation and Model Fusion

Algorithm 1

Logic diagram of the model fusion algorithm.

	Model input:
	Training set D = {(x1, y1), (x2, y2), …, (xm, ym)}
	Primary learner £1, £2, …, £T
	Secondary learner £
	Process:
	1: for t = 1, 2, …, T do
	2:   ht  =  £T (D)
	3: end for
	4: D = Ø;
	5: for t = 1, 2, …, m do
	6:   for t = 1, 2, …, T do
	7:    Zit = ht (xi);
	8:   end for
	9:   Dj  =  DjU ((zi1, zi2, …, ziT), yi)
	10:   hj  =  (Dj)
	11: end for
	Output: H (x) = hj (h1 (x), h2 (x), h3 (x), …, hT (x))