On the convergence of loss and uncertainty-based active learning algorithms

Open in new window