Fig. 9
From: Microalgae classification using semi-supervised and active learning based on Gaussian mixture models
Comparative of semi-supervised learning, in black color, against active learning using entropy, in red color. It is important that the semi-supervised learning be used as initial step to the active learning. Therefore, the semi-supervised results is s linked to the active learning by a blue line. The results obtained using SVM method are trained from the same supervised data used in semi-supervised approach, in green color. The mean and standard deviation are estimated and illustrated in the figure. Two different metrics are evaluated: accuracy and maxF1. a Accuracy comparative in the first dataset. b MaxF1 comparative in the first dataset. c Accuracy comparative in the second dataset. d MaxF1 comparative in the second dataset (color figure online)