Showing 1–1 of 1 results for author: Nishii, R

Minimizing the expected value of the asymmetric loss and an inequality of the variance of the loss

Authors: Naoya Yamaguchi, Yuka Yamaguchi, Ryuei Nishii

Abstract: For some estimations and predictions, we solve minimization problems with asymmetric loss functions. Usually, we estimate the coefficient of regression for these problems. In this paper, we do not make such the estimation, but rather give a solution by correcting any predictions so that the prediction error follows a general normal distribution. In our method, we can not only minimize the expected… ▽ More For some estimations and predictions, we solve minimization problems with asymmetric loss functions. Usually, we estimate the coefficient of regression for these problems. In this paper, we do not make such the estimation, but rather give a solution by correcting any predictions so that the prediction error follows a general normal distribution. In our method, we can not only minimize the expected value of the asymmetric loss, but also lower the variance of the loss. △ Less

Submitted 17 July, 2019; originally announced July 2019.

MSC Class: 62A99; 62E99