基于自然邻居的标记分布学习

doi:10.11896/jsjkx.190700012

计算机科学 ›› 2020, Vol. 47 ›› Issue (8): 132-136.doi: 10.11896/jsjkx.190700012

所属专题：大数据&数据科学虚拟专题

• 数据库&大数据&数据科学 • 上一篇下一篇

基于自然邻居的标记分布学习

姚成亮, 朱庆生

重庆大学计算机学院软件理论与技术重庆市重点实验室重庆 400044

出版日期:2020-08-15 发布日期:2020-08-10
通讯作者: 朱庆生(qszhu@cqu.edu.cn)
作者简介:ycliang@cqu.edu.cn
基金资助:
国家自然科学基金(61802360);重庆市科技项目(KJZH7104)

Label Distribution Learning Based on Natural Neighbors

YAO Cheng-liang, ZHU Qing-sheng

Chongqing Key Lab of Software Theory and Technology, College of Computer Science, Chongqing University, Chongqing 400044, China

Online:2020-08-15 Published:2020-08-10
About author:YAO Cheng-liang, born in 1990, postgraduate.His main research interests include data mining and machine lear-ning.
ZHU Qing-sheng, born in 1956, professor, Ph.D supervisor, is a member of China Computer Federation.His main research interests include service-oriented software engineering, data mining and outlier detection, and machine learning.
Supported by:
This work was supported by the National Natural Science Foundation of China (61802360) and Science and Technology Project of Chongqing (KJZH7104).

摘要/Abstract

摘要： 标记分布是一种新的机器学习范式, 能很好地解决某些标记多义性问题, 可看作多标记的泛化。传统的单标记学习和多标记学习均可看作标记分布学习的特例。已有的标记分布学习算法中, 基于算法改造的AA-KNN(Algorithm Adaptation-KNN)是一种高效的算法, 但任何涉及K近邻求解问题的算法在处理不同数据集时, 参数K值的选取都是一个难题, 不同的K值得到的结果明显不同。基于此, 将自然最近邻居的概念引入标记分布学习, 提出一种新的标记分布学习方法。对数据集使用自然最近邻居搜索算法查找每个样本的自然邻居, 取自然邻居的标记分布均值作为预测结果。搜索算法不需要人工设置任何参数, 同时搜索算法是一种被动搜索, 其自适应计算得到每个样本的邻居。在6个数据集上使用6个评价指标进行实验, 结果表明, 与AA-KNN相比, 结合自然最近邻居的标记分布学习算法不仅避免了人工设置参数的问题, 而且取得了更优的效果。

关键词: 标记分布, 标记分布学习, 无参数, 自然邻居

Abstract: Label distribution is a new machine learning paradigm which can solve some markup ambiguity problems well.Label distribution can be seen as the generalization of multi-label.Traditional single-label learning and multi-label learning can be seen as special cases of label distribution learning.AA-KNN based on algorithm adaptive is an effective algorithm, but it is diffcult to choose an appropriate parameter K which affects the perfomence when KNN is used.So, Natural neighbors is introduced into LDL and a new label distribution learning algorithm is proposed.It finds natural neighbors of each object by searching algorithm, and then gets the average of labels of these neighbours as the predicted result.The natural neighbours searching algorithm does not need any parameter and is passive so that neighbors of each object is decided automatically.Experiments was conducted on 6 data sets and 6 evaluation indexes.The experiments show that the proposed algorithm not only solves the problem of choosing parameter K, but also improves the performance compared with AA-KNN.

Key words: Label distribution, Label distribution learning, Natural neighbors, No-parameter

中图分类号:

TP391

姚成亮, 朱庆生. 基于自然邻居的标记分布学习[J]. 计算机科学, 2020, 47(8): 132-136. https://doi.org/10.11896/jsjkx.190700012

YAO Cheng-liang, ZHU Qing-sheng. Label Distribution Learning Based on Natural Neighbors[J]. Computer Science, 2020, 47(8): 132-136. https://doi.org/10.11896/jsjkx.190700012

参考文献

[1]TSOUMAKAS G, KATAKIS I.Multi-label classification:an overview[J].International Journal of Data Warehousing and Mi-ning, 2007, 3(3):1-13.
[2]GENG X, LUO L.Multilabel ranking with inconsistent rankers[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition.Columbus, 2014:3742-3747.
[3]GENG X.Label distribution learning[J].IEEE Trans KnowlData Eng, 2016, 28:1734-1748.
[4]GENG X, XU N.Label distribution learning and label enhancement[J].Sci Sin Inform, 2018, 48(5):521-530.
[5]ZHANG Z, WANG M, GENG X.Crowd counting in public video surveillance by label distribution learning[J].Neurocomputing, 2015, 166(C):151-163.
[6]GENG X, XIA Y.Head pose estimation based on multivariate label distribution[C]∥Proceedings of IEEE International Conference on Computer Vision and Pattern Recognition, Columbus, USA, 2014:1837-1842.
[7]GENG X, HOU P.Pre-relesse prediction of crowd opinion on movies by label distribution learning[C]∥Proceedings of the International Joint Conference on Artificial Intelligence.Bueons Aires, Argentina, San Francisco, USA:Morgan Kaufmann, 2015:3511-3517.
[8]GENG X, YIN C, ZHOU Z H.Facial age estimation by learning from label distribution[J].IEEE Trasaction on Pattern Analysis and Machine Intelligence, 2013, 35(10):2401-2412.
[9]LING M, GENG X.Indoor Crowd Counting by Mixture ofGaussians Label Distribution Learning[J].IEEE Transactions on Image Processing, 2019, 28(11):5691-5701.
[10]XU L, CHEN J, GAN Y.Head pose estimation using improved label distribution learning with fewer annotations. Multimedia Tools and Applications, 2019, 78(14):19141-19162.
[11]HUANG J, ZHU Q, YANG L, et al.A non-parameter outlier detection algorithm based on Natural Neighbor[J].Knowledge-Based Systems, 2016, 92:71-77.
[12]YANG L, ZHU Q, HUANG J, et al.Adaptive edited naturalneighbor algorithm[J].Neurocomputing, 2017, 230:427-433.
[13]ZHAO Q, GENG X.Selection of target function in label distribution learning[J].Journal of Frontiers of Computer Science and Technology, 2017, 11(5):708-719.
[14]ZHU Q, FENG J, HUANG J.Natural neighbor:A self-adaptive neighborhood method without parameter K[J].Pattern Recognition Letters, 2016, 80(1):30-36.
[15]LIU R, WANG H, YU X.Shared-nearest-neighbor-based clustering by fast search and find of density peaks[J].Information Sciences, 2018, 450:200-226.
[16]ZHU Q, FENG J, HUANG J.Natural neighbor:A self-adaptive neighborhood method without parameter K[J].Pattern Recognition Letters, 2016, 80:30-36.
[17]ZHANG M L, ZHOU Z H.A Review on Multi-Label Learning Algorithms[J].IEEE Transactions on Knowledge & Data Engineering, 2014, 26(8):1819-1837.
[18]CHEN M, LI L J, WANG B, et al.Effectively clustering by finding density backbone based-on kNN[J].Pattern Recognition, 2016, 60:486-4988.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

基于自然邻居的标记分布学习

Label Distribution Learning Based on Natural Neighbors

PDF (PC)

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 6

Metrics

本文评价

推荐阅读 0

[1]	杨旭华, 王磊, 叶蕾, 张端, 周艳波, 龙海霞. 基于节点相似性和网络嵌入的复杂网络社区发现算法 Complex Network Community Detection Algorithm Based on Node Similarity and Network Embedding 计算机科学, 2022, 49(3): 121-128. https://doi.org/10.11896/jsjkx.210200009
[2]	董心悦, 范瑞东, 侯臣平. 基于边际概率分布匹配的主动标记分布学习 Active Label Distribution Learning Based on Marginal Probability Distribution Matching 计算机科学, 2020, 47(9): 190-197. https://doi.org/10.11896/jsjkx.200700077
[3]	徐洋, 陈燚, 黄磊, 谢晓尧. 基于多层BP神经网络和无参数微调的人群计数方法 Crowd Counting Method Based on Multilayer BP Neural Networks and Non-parameter Tuning 计算机科学, 2018, 45(10): 235-239. https://doi.org/10.11896／j.issn.1002-137X.2018.10.043
[4]	朱庆生,段浪军,杨力军. 基于自然邻居和最小生成树的原型选择算法 Prototype Selection Algorithm Based on Natural Neighbor and MST 计算机科学, 2017, 44(4): 241-245. https://doi.org/10.11896/j.issn.1002-137X.2017.04.051
[5]	冯骥,张程,朱庆生. 一种具有动态邻域特点的自适应最近邻居算法 Adaptive Nearest Neighbor Algorithm with Dynamic Neighborhood 计算机科学, 2017, 44(12): 194-201. https://doi.org/10.11896/j.issn.1002-137X.2017.12.036
[6]	龚劬,许凯强. 有监督的无参数核局部保持投影及人脸识别 Parameter-less Supervised Kernel Locality Preserving Projection and Face Recognition 计算机科学, 2016, 43(9): 301-304. https://doi.org/10.11896/j.issn.1002-137X.2016.09.060