Score-Oriented Loss (SOL) functions
作者：
Highlights：
• In deep learning loss minimization and score maximization are intertwined issues.
• SOL functions guarantee an a priori optimization of a given skill score.
• SOL perspective is to treat the score-related threshold as a random variable.
• The optimal threshold is driven by the density map of the a priori distribution.
• Classification tests confirm the automatic threshold optimization provided by SOLs.
摘要
•In deep learning loss minimization and score maximization are intertwined issues.•SOL functions guarantee an a priori optimization of a given skill score.•SOL perspective is to treat the score-related threshold as a random variable.•The optimal threshold is driven by the density map of the a priori distribution.•Classification tests confirm the automatic threshold optimization provided by SOLs.
论文关键词：Supervised machine learning,Binary classification,Loss functions,Skill scores
论文评审过程：Received 26 April 2021, Revised 27 May 2022, Accepted 18 July 2022, Available online 21 July 2022, Version of Record 8 August 2022.
论文官网地址：https://doi.org/10.1016/j.patcog.2022.108913
