论文标题
确保深度学习竞赛中的可重复性
Guaranteeing Reproducibility in Deep Learning Competitions
论文作者
论文摘要
为了鼓励通过可再现和健壮的训练行为开发方法，我们提出了一个挑战范式，即直接评估竞争者的学习程序，而不是预先训练的药物。由于竞争组织者在受控的环境中重新训练了拟议的方法，因此可以保证可重复性，并且 - 通过使用固定的测试集对提交进行重新培训，可以帮助确保概括受过培训的环境。
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.
