- Multi-view learning focuses on how to improve model performance using multiple learners.
→ \to → 避免 how to 这种口语用法
Multi-view learning focuses on using multiple learners to improve model performance. - For semi-supervised regression (SSR), multi-view learners can explore more hidden or less observable information presented by unlabeled data.
→ \to → 如果没有在摘要里面多次出现,就不需要缩写
→ \to → more 在这里表示更加,还是更多 (与其它方案相比)?
→ \to → more 与 less 在这里形成了非常具有误导性的对应,可能你这里的 more 是想把 “hidden” 与 “less observable” 同时修饰
For semi-supervised regression, multi-view learners can explore more hidden information presented by unlabeled data. - However, some studies have found that some low-quality instances and pseudo-labels may degrade the quality of the final regressor.
→ \to → some 出现两次
However, extensive studies have revealed that some unrepresentative instances and low-quality pseudo-labels may degrade the quality of the regressor. - In this paper, we propose the multi-view regression through serial updating and safe labeling (MESS) to face such a fundamental problem.
→ \to → face 表示我在森林里遇到熊的时候,直面它可以死得更光荣
→ \to → address, handle 是常用词
In this paper, we propose the multi-view regression through serial updating and safe labeling (MESS) to face such a fundamental problem.
→ \to → such 是表示一类,与 such as 有一点关联
→ \to → 提出的一个模型还是算法?
→ \to → problem 一般指良好定义的问题 (输入、输出、优化目标、约束条件);issue 指一般性的研究问题
→ \to → 一般使用 address … issue, handle … problem (不敢确信)
In this paper, we propose a multi-view regression through serial updating and safe labeling (MESS) model to address this fundamental issue. - First, the serial update technique is used for the validity of selected unlabeled instances.
→ \to → 先把句子减短: Techqniue is used for validity.
First, the serial update technique is used for improving the validity of selected unlabeled instances.
First, the serial update technique improves the validity of selected unlabeled instances. - This strategy is fully compatible with the safe regression framework and brings a better label quality than the traditional parallel way.
→ \to → 尽可能不要用 and, 除非有非常良好的对仗
Since this strategy is fully compatible with the safe regression framework, it brings a better label quality than the traditional parallel way.
前者可控,后者是效果
论文修改建议 (Y. Li 20210922)
最新推荐文章于 2024-05-20 20:08:58 发布