当前位置: X-MOL 学术Test › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Rejoinder on: “On active learning methods for manifold data”
TEST ( IF 1.2 ) Pub Date : 2020-01-02 , DOI: 10.1007/s11749-019-00697-9
Hang Li , Enrique Del Castillo , George Runger

We thank the discussants for their comments and careful reading of our manuscript, which have enhanced and complemented our presentation. We also thank the editors of TEST for this opportunity to clarify some aspects of our work in more detail. In what follows, we first address some points touched by both sets of discussants, and then consider comments made individually by each of them. We conclude with a description of a method that can improve the speed of the retraining required in the SSGP-AL method when used for classification by re-using previous learning as opposed to re-estimating the GP model from scratch at each AL cycle.

中文翻译:

重新加入:“关于多数据的主动学习方法”

我们感谢讨论者的评论和对手稿的认真阅读,这些内容对我们的演讲有所补充和补充。我们也感谢TEST的编辑提供的这次机会,以便更详细地阐明我们工作的某些方面。接下来,我们首先讨论两组讨论者都涉及的一些问题,然后再考虑它们各自的评论。我们以对方法的描述作为结束,该方法可以通过重新使用先前的学习来提高SSGP-AL方法用于分类时所需的再训练速度,而不是在每个AL周期从头开始重新估计GP模型。
更新日期:2020-01-02
down
wechat
bug