Augmentation by Counterfactual Explanation - Fixing an Overconfident Classifier.

A highly accurate but overconfident model is ill-suited for deployment in critical applications such as healthcare and autonomous driving. The classification outcome should reflect a high uncertainty on ambiguous in-distribution samples that lie close to the decision boundary. The model should also refrain from making overconfident decisions on samples that lie far outside its training distribution, far-out-of-distribution (far-OOD), or on unseen samples from novel classes that lie near its training distribution (near-OOD).

'Very brief advice' (VBA) on smoking in family practice: a qualitative evaluation of the tobacco user's perspective.

Background: Very Brief Advice on smoking (VBA) is an evidence-based intervention designed to increase quit attempts among patients who smoke. VBA has been widely disseminated in general practice settings in the United Kingdom, however its transferability to Southern European settings is not well established. This study sought to document the perspectives of Greek general practice patients in terms of the acceptability and satisfaction with receiving VBA from their general practitioner (GP) and its influence on patients' motivation to make a quit attempt.