ID 原文 译文
763 实验结果表明本文所提出的图像修复算法性能优于多种现有算法。 Experimental results show that the proposed approach outperforms several state-of-the-art methods.
764 Abstract-Refine(抽象—精炼)方法是软件模型检测领域中较为有效的设计思想，具有较高的通用性和效率优势， Abstract-Refine is a relatively effective method in the field of software model checking, which has the ad-vantages of high generality and efficiency.
765 但目前并没有一个框架可以对其精确进行描述及实现有效的模块化使用和替换。 However, there is no framework for precise description and effective modular useor replacement of this method so far.
766 本文提出了一种模块化的Abstract-Refine 算法框架，分析和解释了 Abstract-Refine 算法所接受的输入程序的精细结构和特性， This paper introduces a modular Abstract-Refine algorithm framework which analyzesand explains the structure of input program in fine-grained level.
767 并对 Abstract-Re-fine 算法和相关子算法运用平衡操作符做以模块化解耦，使得子算法的修改和更换不需要依赖对上层的变更。 Also, this method modularly decouples Abstract-Refine al-gorithm from its sub-algorithms with the balancing operator, so that any modifications on sub-algorithms will not affect the upper level.
768 经过实验验证，本方法可有效实现传统算法模块化解耦，同时不对原算法的性能造成冲击。 Experiments verify that our approach can effectively implement modular decoupling of traditional algorithms, and will not impact the performance of original algorithms.
769 多目标回归学习是指同时学习多个相关的回归任务，其主要挑战来自于对输入要素和输出目标变量之间的基础关系进行建模以及对目标间的相关性进行探索。 Multi-target regression (MTR)refers to learning multiple relevant regression tasks simultaneously. The main challenges of multi-target regression arise from modeling the underlying relationships between input features and outputtarget variables as well as exploring inter-target correlations.
770 针对这两个挑战，本文提出了一种基于标签特定特征的多目标回归稀疏集成方法，通过探索目标间的相关性，为每个目标构建其独特的标签特定特征，提高算法整体的预测精度; In this paper, we propose a multi-target regression method viasparse integration and label-specific features (SI-LSF)that utilizes inter-target correlations to improve the overall prediction accuracy by constructing label-specific features and deals with the input-output relationships through sparse integration of va-rious regression models.
771 同时设计一种稀疏性聚合函数对不同的回归方法进行集成，从而处理输入与输出间的复杂关系。在 18 个数据集上与有代表性的多目标回归方法进行对比实验，充分证明了本文方法的有效性与竞争性。 Extensive experimental evaluation on 18 benchmark datasets demonstrates that our proposed method can achieve competitive performance against representative state-of-the-art multi-target regression methods, which shows the great effectiveness in dealing with multivariate prediction.
772 大数据分布式存储系统中，修复流水线(Repair Pipelining，RP)减少 90% 的修复时间，有效地解决由于修复时间开销较大，纠删码不适用于存储热数据的问题。 In distributed storage system for big data, the repair pipelining (RP)reduces repair time by 90% , whicheffectively solves the problem that erasure code is not suitable for storing hot data due to the heavy overhead of repair time.