On Mixup Training: Improved Calibration and
    Predictive Uncertainty for Deep Neural Networks

            Sunil Thulasidasan,1,2 , Gopinath Chennupati1 , Jeff Bilmes2 ,
                Tanmoy Bhattacharya1 , Sarah Michalak1
                   1
                    Los Alamos National Laboratory
       2
         Department of Electrical and Computer Engineering, University of Washington

                         Abstract
       Mixup [37] is a recently proposed method for training deep neura ...