确保深度学习竞赛的再现性
Guaranteeing Reproducibility in Deep Learning Competitions
摘要:
为了鼓励开发具有可重复和强大训练行为的方法,我们提出了一种挑战范式,在这种范式中,竞争对手直接根据其学习过程的性能而不是预先训练的代理进行评估。由于竞赛组织者在一个受控的环境中重新训练提出的方法,他们可以保证可重复性,并且-通过使用一个保持的测试集重新训练提交-帮助确保泛化过去的环境,他们被训练。
英文摘要:
To encourage the development of methods with reproducible and robust training behavior, we propose a challenge paradigm where competitors are evaluated directly on the performance of their learning procedures rather than pre-trained agents. Since competition organizers re-train proposed methods in a controlled setting they can guarantee reproducibility, and -- by retraining submissions using a held-out test set -- help ensure generalization past the environments on which they were trained.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。