大师作文网英译汉（要求语句通顺）
来源：学生作业帮 编辑：大师作文网作业帮 分类：英语作业 时间：2024/09/29 10:21:17
英译汉（要求语句通顺）
We proposed online-learning algorithms for learning diver- sity in rankings.The proposed algorithms balance diversity and relevance by modeling the utility of the ranking as a submodular function.Using plausible user feedback in the form of preferences between rankings,the algorithms are able to learn rankings that optimize the user’s utility.In addition to theoretically characterizing the performance of the algorithms and their robustness to noise,we showed that the algorithms perform well in empirical studies.Future re- search directions are the deployment of the algorithm in a
real system and validation of the feedback model in user studies.This research was funded in part by NSF Award IIS- 0905467.
We proposed online-learning algorithms for learning diver- sity in rankings.The proposed algorithms balance diversity and relevance by modeling the utility of the ranking as a submodular function.Using plausible user feedback in the form of preferences between rankings,the algorithms are able to learn rankings that optimize the user’s utility.In addition to theoretically characterizing the performance of the algorithms and their robustness to noise,we showed that the algorithms perform well in empirical studies.Future re- search directions are the deployment of the algorithm in a
real system and validation of the feedback model in user studies.This research was funded in part by NSF Award IIS- 0905467.
针对不同级别的学习过程,我们提出了在线学习算法.该算法以各级别的效用为次模函数建模,从而平衡了各级别的多样性和相关性.通过搜集用户关于不同级别偏好的反馈信息,该算法得以确定最优级别,使用户效用最大化.除了对该算法的特点和稳健性进行理论说明,我们还在实证研究中验证了该算法.未来的研究方向将关注该算法在真实系统中的运用,以及对用户研究中的反馈模型进行验证.本研究由美国国家科学基金会奖项IIS-0905467提供部分资助.
【非计算机专业人士,专业术语准确性欠佳,见谅.】
【非计算机专业人士,专业术语准确性欠佳,见谅.】