Developing Learning Algorithms via Optimized Discretization of Continuous Dynamical Systems
Tao, Qing1,2; Sun, Zhengya1; Kong, Kang2
2012-02-01
发表期刊IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS
卷号42期号:1页码:140-149
文章类型Article
摘要Most of the existing numerical optimization methods are based upon a discretization of some ordinary differential equations. In order to solve some convex and smooth optimization problems coming from machine learning, in this paper, we develop efficient batch and online algorithms based on a new principle, i.e., the optimized discretization of continuous dynamical systems (ODCDSs). First, a batch learning projected gradient dynamical system with Lyapunov's stability and monotonic property is introduced, and its dynamical behavior guarantees the accuracy of discretization-based optimizer and applicability of line search strategy. Furthermore, under fair assumptions, a new online learning algorithm achieving regret O(root T) or O(logT) is obtained. By using the line search strategy, the proposed batch learning ODCDS exhibits insensitivity to the step sizes and faster decrease. With only a small number of line search steps, the proposed stochastic algorithm shows sufficient stability and approximate optimality. Experimental results demonstrate the correctness of our theoretical analysis and efficiency of our algorithms.
关键词Dynamical Systems Line Search Machine Learning Online Learning Optimization Algorithms Projected Subgradient Algorithms Regret
WOS标题词Science & Technology ; Technology
关键词[WOS]NEURAL-NETWORK
收录类别SCI
语种英语
WOS研究方向Automation & Control Systems ; Computer Science
WOS类目Automation & Control Systems ; Computer Science, Artificial Intelligence ; Computer Science, Cybernetics
WOS记录号WOS:000302096700011
引用统计
被引频次:1[WOS]   [WOS记录]     [WOS相关记录]
文献类型期刊论文
条目标识符http://ir.ia.ac.cn/handle/173211/10743
专题数字内容技术与服务研究中心_听觉模型与认知计算
作者单位1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China
2.New Star Res Inst Appl Technol, Hefei 230031, Peoples R China
推荐引用方式
GB/T 7714
Tao, Qing,Sun, Zhengya,Kong, Kang. Developing Learning Algorithms via Optimized Discretization of Continuous Dynamical Systems[J]. IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS,2012,42(1):140-149.
APA Tao, Qing,Sun, Zhengya,&Kong, Kang.(2012).Developing Learning Algorithms via Optimized Discretization of Continuous Dynamical Systems.IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS,42(1),140-149.
MLA Tao, Qing,et al."Developing Learning Algorithms via Optimized Discretization of Continuous Dynamical Systems".IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS 42.1(2012):140-149.
条目包含的文件
条目无相关文件。
个性服务
推荐该条目
保存到收藏夹
查看访问统计
导出为Endnote文件
谷歌学术
谷歌学术中相似的文章
[Tao, Qing]的文章
[Sun, Zhengya]的文章
[Kong, Kang]的文章
百度学术
百度学术中相似的文章
[Tao, Qing]的文章
[Sun, Zhengya]的文章
[Kong, Kang]的文章
必应学术
必应学术中相似的文章
[Tao, Qing]的文章
[Sun, Zhengya]的文章
[Kong, Kang]的文章
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。