一种NMI结合HSIC0的台风移动轨迹多因素相关分析方法

doi:10.12068/j.issn.1005-3026.2023.09.003

东北大学学报（自然科学版） ›› 2023, Vol. 44 ›› Issue (9): 1234-1244.DOI: 10.12068/j.issn.1005-3026.2023.09.003

一种NMI结合HSIC0的台风移动轨迹多因素相关分析方法

乔百友¹，郝元卿¹，唐忠¹，汪锐²

(1. 东北大学计算机科学与工程学院，辽宁沈阳110819； 2. 中国科学院沈阳自动化研究所，辽宁沈阳110169)

发布日期:2023-09-28
通讯作者: 乔百友
作者简介:乔百友(1970-)，男，甘肃礼县人，东北大学副教授.
基金资助:
国家重点研发计划项目(2019YFB1405302)；国家自然科学基金资助项目(61872072).

A Multi-factor Correlation Analysis Method for Typhoon Moving Track Based on NMI and HSIC0

QIAO Bai-you¹， HAO Yuan-qing¹， TANG Zhong¹， WANG Rui²

1. School of Computer Science & Engineering， Northeastern University， Shenyang 110819， China； 2. Shenyang Institute of Automation， Chinese Academy of Sciences， Shenyang 110169， China.

Published:2023-09-28
Contact: QIAO Bai-you
About author:-
Supported by:
-

摘要/Abstract

摘要： 现有非线性相关分析方法准确性低、计算代价大，因而不适合大规模、高维度台风轨迹数据相关分析.针对这一问题，首次将希尔伯特-施密特独立准则经验估计(Hilbert-Schmidt independent criterion empirical estimation， HSIC0)引入到台风移动轨迹相关研究中，提出了一种基于标准化互信息(normalized mutual information，NMI)结合HSIC0的多因素相关分析方法.该方法首先利用NMI来过滤掉台风数据中相关性低的冗余因素，然后采用XGBoost机器学习模型来剔除掉无效因素，从而降低后续计算代价.在此基础上，采用基于HSIC0的多因素相关分析方法对台风数据进行了分析，挖掘出了相关性较强的台风移动轨迹影响因素组合，从而提高了台风移动轨迹的预测精度.真实台风数据集上的一系列实验结果表明，提出的方法在MSE，MAE，R^○2指标上均优于NMI、Pearson相关系数、距离相关系数等分析方法.

关键词: 台风移动轨迹；相关分析；多因素；HSIC0；XGBoost

Abstract: The existing nonlinear correlation analysis methods have low accuracy and high computational cost， making them unsuitable for the correlation analysis of large-scale and high-dimensional typhoon track data.To solve this problem，the Hilbert-Schmidt independent criterion empirical estimation (HSIC0) is introduced into typhoon track correlation study for the first time， and a multi-factor correlation analysis method based on normalized mutual information (NMI) and HSIC0 is proposed. The method first uses NMI to filter out redundant factors with low correlation in typhoon data， and then uses XGBoost to eliminate invalid factors， thus reducing the subsequent computational costs.On this basis， a multi-factor correlation analysis method based on HSIC0 is used to analyze typhoon data， and a combination of factors affecting typhoon moving track with strong correlation is mined， thereby improving the prediction accuracy of typhoon moving track.A series of experimental results on real typhoon data sets show that the proposed method outperforms the correlation analysis methods such as NMI， Pearson correlation coefficient， and distance correlation coefficient in indicators such as MSE， MAE， R^○2.

Key words: typhoon moving track; correlation analysis; multi-factor; HSIC0; XGBoost

中图分类号:

TP301
P444

乔百友，郝元卿，唐忠，汪锐. 一种NMI结合HSIC0的台风移动轨迹多因素相关分析方法[J]. 东北大学学报（自然科学版）, 2023, 44(9): 1234-1244.

QIAO Bai-you， HAO Yuan-qing， TANG Zhong， WANG Rui. A Multi-factor Correlation Analysis Method for Typhoon Moving Track Based on NMI and HSIC0[J]. Journal of Northeastern University(Natural Science), 2023, 44(9): 1234-1244.

参考文献

[1]Zhang L，Zhu H，Liu J.Characteristics of tropical cyclones formed in the Eastern Pacific Northwest［J］.Natural Hazards，2021，106(3):2619-2633.
[2]Huang X Y，Jin L.An artificial intelligence prediction model based on principal component analysis for typhoon tracks［J］.Chinese Journal of Atmospheric Sciences，2013，37(5):1154-1164.
[3]Vazirani V V.Approximation algorithms［M］.Berlin:Springer-Verlag，2001.
[4]Mayer-Schonberger V，Cukier K.Big data:a revolution that will transform how we live，work，and think［M］.New York:Eamon Dolan/Houghton Mifflin Harcourt，2013.
[5]黄奕武，高拴柱，钱奇峰.T639台风预报误差与环境场变量的相关分析和回归分析［J］.气象，2016，42(12):1506-1512.(Huang Yi-wu，Gao Shuan-zhu，Qian Qi-feng.Correlation and regression analysis of typhoon forecast errors and ambient variables by T639［J］.Meteorological Monthly，2016，42(12):1506-1512.)
[6]Reshef D N，Reshef Y A，Finucane H K，et al.Detecting novel associations in large data sets［J］.Science，2011，334(6062):1518-1524.
[7]Székely G J，Rizzo M L，Bakirov N K.Measuring and testing dependence by correlation of distances［J］.The Annals of Statistics，2007，35(6):2769-2794.
[8]Gretton A，Bousquet O，Smola A，et al.Measuring statistical dependence with Hilbert-Schmidt norms［C］// Proceedings of the 16th International Conference on Algorithmic Learning Theory.Singapore:Sanjay Jain，2005:63-77.
[9]Chen X，Duan Y，Houthooft R，et al.InfoGAN:interpretable representation learning by information maximizing generative adversarial nets［C］//The 30th Annual Conference on Neural Information Processing Systems.Barcelona，2016:2172-2180.
[10]胡兵兵，唐华，吴幼龙.基于互信息约束的生成对抗网络分类模型［J］.中国科学院大学学报，2022，39(4):551-560.(Hu Bing-bing，Tang Hua，Wu You-long.Classification models based on generative adversarial networks with mutual information regularization［J］.Journal of University of Chinese Academy of Sciences，2022，39(4):551-560.)
[11]Vinh L T，Lee S，Park Y T，et al.A novel feature selection method based on normalized mutual information［J］.Applied Intelligence，2012，37(1):100-120.
[12]王黎明，吴香华，赵天良，等.基于距离相关系数和支持向量机回归的PM_2.5浓度滚动统计预报方案［J］.环境科学学报，2017，37(4):1268-1276.(Wang Li-ming，Wu Xiang-hua，Zhao Tian-liang，et al.A scheme for rolling statistical forecasting of PM_2.5 concentrations based on distance correlation coefficient and support vector regression［J］.Acta Scientiae Circumstantiae，2017，37(4):1268-1276.)
[13]Wen C，Yang Y，Xiao Q，et al.Genome-wide association studies of brain imaging data via weighted distance correlation［J］.Bioinformatics，2020，36(19):4942-4950.
[14]梁吉业，冯晨娇，宋鹏.大数据相关分析综述［J］.计算机学报，2016，39(1):1-18.(Liang Ji-ye，Feng Chen-jiao，Song Peng.A survey on correlation analysis of big data［J］.Chinese Journal of Computers，2016，39(1):1-18.)
[15]宋成宝，柳平增，刘兴华，等.基于HSIC的日光温室温度传感器优化配置策略［J］.农业工程学报，2022，38(8):200-208.(Song Cheng-bao，Liu Ping-zeng，Liu Xing-hua，et al.Optimal configuration strategy for temperature sensors in solar greenhouse based on HSIC［J］.Transactions of the Chinese Society of Agricultural Engineering，2022，38(8):200-208.)
[16]李柏松，任伟杰，韩敏.基于HSIC-GL的多元时间序列非线性Granger因果关系分析［J］.信息与控制，2021，50(3):356-365.(Li Bai-song，Ren Wei-jie，Han Min.Nonlinear Granger causality analysis for multivariate time series using HSIC-GL model［J］.Information and Control，2021，50(3):356-365.)
[17]张晓琴，刘玲，郭鑫垚.基于HSIC0的类间非线性相关系数度量［J］.计算机工程与应用，2019，55(3):46-49.(Zhang Xiao-qin，Liu Ling，Guo Xin-yao.Measurement of nonlinear correlation coefficient among classes based on HSIC0［J］.Computer Engineering and Applications，2019，55(3):46-49.)
[18]Pearson K.Contributions to the mathematical theory of evolution(III):regression，heredity，and panmixia［J］.Proceedings of the Royal Society of London，1895，59(1):69-71.
[19]Chen T Q，Guestrin C.XGBoost:a scalable tree boosting system［C］// Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining.New York，2016:785-794.
[20]Song L，Smola A J，Gretton A，et al.Feature selection via dependence maximization［J］.Journal of Machine Learning Research，2012，13:1393-1434.
[21]Ying M，W Zhang，H Yu，et al.An overview of the China Meteorological Administration tropical cyclone database［J］.Journal of Atmospheric and Oceanic Technology，2014，31(2):287-301.
[22]Lu X，Yu H，Ying M.et al.Western north pacific tropical cyclone database created by the China Meteorological Administration［J］.Advances in Atmospheric Sciences，2021，38(4):690-699.
[23]Kalnay E，Kanamitsu M，Kistler R，et al.The NCEP/NCAR 40-year reanalysis project［J］.Bulletin of the American Meteorological Society，1996，77(3):437-470.
[24]National Oceanic and Atmospheric Administration，National Center for Atmospheric Research.NCEP reanalysis datasets (1948-2018).A big earth data platform for three poles［EB/OL］.［2022-03-20］.https://doi.org/10.11888/Meteoro.tpdc.270922.

一种NMI结合HSIC0的台风移动轨迹多因素相关分析方法

A Multi-factor Correlation Analysis Method for Typhoon Moving Track Based on NMI and HSIC0

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 0

编辑推荐

Metrics

本文评价