• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>Neural Networks and Learning Systems, IEEE Transactions on >Scalable Kernel Ordinal Regression via Doubly Stochastic Gradients
【24h】

Scalable Kernel Ordinal Regression via Doubly Stochastic Gradients

机译：通过双随机梯度可扩展内核序列回归

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Ordinal regression (OR) is one of the most important machine learning tasks. The kernel method is a major technique to achieve nonlinear OR. However, traditional kernel OR solvers are inefficient due to increased complexity introduced by multiple ordinal thresholds as well as the cost of kernel computation. Doubly stochastic gradient (DSG) is a very efficient and scalable kernel learning algorithm that combines random feature approximation with stochastic functional optimization. However, the theory and algorithm of DSG can only support optimization tasks within the unique reproducing kernel Hilbert space (RKHS), which is not suitable for OR problems where the multiple ordinal thresholds usually lead to multiple RKHSs. To address this problem, we construct a kernel whose RKHS can contain the decision function with multiple thresholds. Based on this new kernel, we further propose a novel DSG-like algorithm, DSGOR. In each iteration of DSGOR, we update the decision functional as well as the function bias with appropriately set learning rates for each. Our theoretic analysis shows that DSGOR can achieve O(1/t) convergence rate, which is as good as DSG, even though dealing with a much harder problem. Extensive experimental results demonstrate that our algorithm is much more efficient than traditional kernel OR solvers, especially on large-scale problems.
机译：序数回归（或）是最重要的机器学习任务之一。内核方法是实现非线性的主要技术。然而，由于多个序阈值的复杂性增加以及内核计算的成本增加，传统的内核或溶解度效率低。双随机梯度（DSG）是一种非常有效且可扩展的内核学习算法，其将随机特征近似与随机功能优化相结合。然而，DSG的理论和算法只能支持独特的再现内核Hilbert空间（RKHS）中的优化任务，这不适合或问题，其中多个序阈值通常导致多个RKHSS。要解决此问题，我们构建一个内核，其RKHS可以包含多个阈值的决策功能。基于这一新内核，我们进一步提出了一种新的DSG样算法，DSGOR。在DSGOR的每次迭代中，我们更新决策功能以及具有适当设置每个学习速率的函数偏差。我们的理论分析表明，DSGOR可以实现O（1 / T）收敛速率，这也与DSG一样好，即使处理更难的问题。广泛的实验结果表明，我们的算法比传统的内核或溶剂更有效，尤其是在大规模问题上。

著录项

  • 来源
  • 作者

    Gu Bin; Geng Xiang; Li Xiang; Shi Wanli; Zheng Guansheng; Deng Cheng; Huang Heng;

  • 作者单位

    Nanjing Univ Informat Sci & Technol Jiangsu Engn Ctr Network Monitoring Nanjing 210044 Peoples R China|Nanjing Univ Informat Sci & Technol Sch Comp & Software Nanjing 210044 Peoples R China|JD Finance Amer Corp Mountain View CA USA;

    Nanjing Univ Informat Sci & Technol Sch Comp & Software Nanjing 210044 Peoples R China;

    Univ Western Ontario Comp Sci Dept London ON N6A 3K7 Canada;

    Nanjing Univ Informat Sci & Technol Sch Comp & Software Nanjing 210044 Peoples R China;

    Nanjing Univ Informat Sci & Technol Sch Comp & Software Nanjing 210044 Peoples R China;

    Xidian Univ Sch Elect Engn Xian 710114 Peoples R China;

    JD Finance Amer Corp Mountain View CA USA|Univ Pittsburgh Dept Elect & Comp Engn Pittsburgh PA 15260 USA;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

    Doubly stochastic gradients (DSGs); kernel learning; ordinal regression (OR); random features;

    机译：双随机梯度（DSG）;核心学习;序数回归（或）;随机特征;

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号