(3.227.0.150) 您好!臺灣時間:2021/05/06 13:08
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

我願授權國圖
: 
twitterline
研究生:Dini Rahmawati
研究生(外文):Dini Rahmawati
論文名稱:以時空模型預測臺灣的登革熱疫情
論文名稱(外文):Using Spatiotemporal Model to Predict Dengue Fever Epidemics in Taiwan
指導教授:黃有評黃有評引用關係
口試委員:蘇國和洪茂盛楊棧雲黃正民
口試日期:2016-07-15
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:電機工程系所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
畢業學年度:104
中文關鍵詞:Data miningDengue feverC-support vector classificationPrediction modelParameter optimization.Association rules
外文關鍵詞:登革熱
相關次數:
  • 被引用被引用:0
  • 點閱點閱:209
  • 評分評分:系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔系統版面圖檔
  • 下載下載:64
  • 收藏至我的研究室書目清單書目收藏:1
The outbreak of dengue fever in southern Taiwan puts the whole nation on alert. Climate change risk people’s health condition through local variation influenced by daily precipitation, air temperature and rapidly unplanned urbanization, which advance expansive diseases and boost the widespread of dengue fever. Those attributes may cause dengue fever to risk medical complications or deaths. The shortcoming of expertise and computing resources drive certain numbers of dengue cases are unrecorded and many cases are misclassified. With respect to these intentions, in this thesis an approached from linear and RBF kernel function optimization of C-Support Vector Classification (C-SVC) used as base classifiers to forecast the statistical dengue fever outbreak that has association with location, air temperature, and daily precipitation.
The research objectives to provide up-to-date dengue fever forecast with high accuracy that require few attributes. Thus, the discovery of classification of optimize C-SVC RBF base of spatiotemporal association rules use Information Gain Ratio of attributes to define a unique query of association rule mining. This result used to provide an accurate forecast that will give appropriate initial warning to citizens or government so that the outbreaks of Dengue Fever can be mitigate.
The performance of the framework was analyzed using statistical reported Dengue Fever Case combine with climatological nationwide data of Taiwan. This study considers the nationwide weekly data of maximum temperatures, minimum temperatures, average temperatures (in Celsius degree) and daily precipitation (mm) during the periods of 2014 to 2015 for experiments. This study used nationwide weekly data of indigenous and imported DF cases during the periods of 2014 to 2015 for experiments. These data were obtained from the Central Epidemic Command Center (CECC) for Dengue Outbreak Taiwan from the Center for Disease Control, Ministry of Health and Welfare, Taiwan. Experimental results verify that RBF kernel after parameter optimization achieves better prediction accuracy on the dengue fever outbreak.
The outbreak of dengue fever in southern Taiwan puts the whole nation on alert. Climate change risk people’s health condition through local variation influenced by daily precipitation, air temperature and rapidly unplanned urbanization, which advance expansive diseases and boost the widespread of dengue fever. Those attributes may cause dengue fever to risk medical complications or deaths. The shortcoming of expertise and computing resources drive certain numbers of dengue cases are unrecorded and many cases are misclassified. With respect to these intentions, in this thesis an approached from linear and RBF kernel function optimization of C-Support Vector Classification (C-SVC) used as base classifiers to forecast the statistical dengue fever outbreak that has association with location, air temperature, and daily precipitation.
The research objectives to provide up-to-date dengue fever forecast with high accuracy that require few attributes. Thus, the discovery of classification of optimize C-SVC RBF base of spatiotemporal association rules use Information Gain Ratio of attributes to define a unique query of association rule mining. This result used to provide an accurate forecast that will give appropriate initial warning to citizens or government so that the outbreaks of Dengue Fever can be mitigate.
The performance of the framework was analyzed using statistical reported Dengue Fever Case combine with climatological nationwide data of Taiwan. This study considers the nationwide weekly data of maximum temperatures, minimum temperatures, average temperatures (in Celsius degree) and daily precipitation (mm) during the periods of 2014 to 2015 for experiments. This study used nationwide weekly data of indigenous and imported DF cases during the periods of 2014 to 2015 for experiments. These data were obtained from the Central Epidemic Command Center (CECC) for Dengue Outbreak Taiwan from the Center for Disease Control, Ministry of Health and Welfare, Taiwan. Experimental results verify that RBF kernel after parameter optimization achieves better prediction accuracy on the dengue fever outbreak.
ABSTRACT i
ACKNOWLEDGMENTS iii
Table of Contents v
List of Figures viii
List of Tables ix
Chapter 1 Introduction 1
1.1. Background 1
1.2. Spatiotemporal Predicting Model Concerns 2
1.3. Goal of Spatiotemporal Model 4
1.4. Previous Related Works 5
1.5. Problem Statement 7
1.6. Thesis Development 9
Chapter 2 Literature Review 11
2.1 Introduction 11
2.2 Spatiotemporal Object 12
2.2.1 Spatial Objects 12
2.2.2 Temporal Objects 13
2.3 Data Mining 14
2.3.1 Predictive Modelling 14
2.3.2 Feature Selection 15
2.4 Spatiotemporal Association Rules (STARs) 17
2.4.1 Hot-Spot Region 17
2.4.2 High Traffic Areas and Stationary Regions 18
2.5 Support Vector Classification (C-SVC) 19
2.5.1 Radial Base Function (RBF) Kernel 20
2.5.2 Linear Kernel 20
2.5.3 Optimization Parameter 21
2.5.4 K-Fold Cross Validation 22
Chapter 3 Data Preparation 24
3.1 Introduction 24
3.2 Data Set 25
3.3 Climatology Data 28
3.4 Daily Dengue Fever Case Statistic Data 29
Chapter 4 Methods 30
4.1 Introduction 30
4.2 Proposed Framework 31
4.2.1 Categorization and Handling Missing Value 32
4.2.2 SVM Classifier 33
4.2.3 Query Extraction 34
4.2.4 Association Rule Mining (STARs) 34
4.3 Performance Measures 35
4.4 Mining Spatiotemporal Association Rules 35
Chapter 5 Data Mining 38
5.1 Introduction 38
5.2 Preprocessing 38
5.3 Information Gain Ratio 40
5.4 Spatiotemporal Association Rules (STARs) 41
Chapter 6 Experimental Results 44
6.1 Introduction 44
6.2 Classification Results 44
6.3 Information Gain Ratio Result 48
6.4 System Performances 50
6.4 STARs Evaluation 51
Chapter 7 Conclusions and Future Works 53
7.1 Conclusions 53
7.2 Future Works 53
References 55
About the Author 59
[1]J.-W. Han, M. Kamber and J. Pei, Data Mining; Concept and Techniques, 3rd Ed. Waltham, MA: Morgan Kaufmann, 2012, pp.408.
[2]J. Novakovic and A. Veljovic, “C-support vector classification: selection of kernel and parameter in medical diagnosis,” in Proc. of Symposium on Intelligent Systems and Informatics (SISY), Subotica, Serbia, pp.465-470, September 2011.
[3]C.-C. Chang and C.-J. Lin, “LIBSVM: a library for support vector machines,” ACM Transactions on Intelligent Systems and Technology, vol. 2, no. 3, pp.1-27, April 2011.
[4]B.E. Boser, I.M. Guyon and V.N. Vapnik, “A training algorithm for optimal margin classifiers,” in Proc. of the 5th Annual ACM Workshop on Computational Learning Theory (COLT), Pittsburgh, PA, USA, pp.144-152, July 1992.
[5]J. Cervantes, X. Li and W. Yu, “SVM classification for large data sets by considering models of classes distribution,” in Proc. Sixth Mexican Int. Conf. in Artificial Intelligence (MICAI), Aguascalientes, Mexico, pp.51-60, November 2007.
[6]T. Hastie, R. Tibshirani and J. Friedman, Data Mining, Interference and Prediction, in The Elements of Statistical Learning, Stanford, California, USA, Springer, p.242, 2009.
[7]J. Cervantes, X. Li and W. Yu, “A novel SVM classification method for large data sets,” in Proc. of IEEE Int. Conf. on Granular Computing (GrC), San Jose, CA, USA, pp.297-302, August 2010.
[8]Rapidminer. (1994). Support Vector Machine (LibSVM) [Online]. Available: http://docs.rapidminer.com
[9](2014, November 24). Dengue Fever [Online]. Available: http://www.cdc.gov.tw
[10]M. Award L. Khan, F. Bastani and I.L. Yen, “An effective support vector machine (SVMs) performance using hierarchical clustering,” in Proc. 16th IEEE Int. Conf. on Tools with Artificial Intelligence (ICTAI), Boca Raton, FL, USA, pp.663-667, November 2004.
[11]U. Maulik and D. Chakraborty, “Fuzzy preference based feature selection and semi supervised SVM for cancer classification,” IEEE Trans. on Nanobioscience, vol. 13, no. 2, pp. 152-160, June 2014.
[12]J. Cervantes, X. Li and W. Yu, “A novel SVM classification method for large data sets,” in Proc. of IEEE Int. Conf. on Granular Computing (GrC), San Jose, CA, USA, pp.297-302, August 2010.
[13]J.H. Faghmous, M. Lee, M. Uluyol, V. Kumar and S. Chatterjee, “A parameter-free spatiotemporal pattern mining model to catalog global ocean dynamics,” in Proc. of IEEE Int. Conf on Data Mining, Dallas, TX, USA, pp.151-160, December 2013.
[14]M. Erwig, M. Schnieder and R.H. Guting “Temporal objects for spatiotemporal data models and a comparison of their Representations,” in Proc. of the Workshops on Data Warehousing and Data Mining: Advances in Database Technologies, London, UK, pp.454-465, November 1998.
[15]A. Mohan, A New Spatiotemporal Data Mining Method and its Application to Reservoir System Operation, Thesis, The Graduate Collage University of Nebraska, USA, May 2014.
[16]C.H. Yu, W. Ding, M. Morabito and P. Chen, “Hierarchical spatiotemporal pattern discovery and predictive modeling,” IEEE Trans. on Knowledge and Data Engineering, vol. 28, no. 4, pp.979-993, April 2016.
[17]N. Rachata, P. Charoenkwan, T. Yooyativong, K. Chamnongthal, C. Lursinsap , and K. Higuchi, “Automatic prediction system of dengue hemorrhagic-fever outbreak risk by using entropy and artificial neural network,” in Proc. of Symposium on Communication and Information Technologies (ISCIT), Vientiane, Laos, pp.210-214, October 2008.
[18]H. Alouaoui, S.Y. Turki and S. Faiz, “Mining spatiotemporal associations using queries,” in Proc. of IEEE Int. Conf. on Information Technology and e-Services (ICITeS), Sousse, Tunisia, pp.1-5, March 2012.
[19]F. Verhein and S. Chawla, “Mining spatiotemporal association rules, sources, sinks, stationary regions and thoroughfares in object mobility databases,” in Proc. 11th International Conference Database System for Advanced Applications (DASFAA), Singapore, pp.187-201, April 2006.
[20]A. Balasundaram and P.T.V. Bhuvaneswari, “Comparative study on decision tree based data mining algorithm to assess risk of epidemic,” in Proc. of IET Chennai Fourth Int. Conf. Sustainable Energy and Intelligent Systems (SEISCON), Chennai, India, pp.390-396, December 2013.
[21]C.J. Paciorek , J.D. Yanosky, R.C. Puett, F. Laden and H.H. Suh “Practical large-scale spatiotemporal modeling of particulate matter concentrations,” Trans Institute of Mathematical Statistics in The Annals of Applied Statistics, vol. 3, No. 1, pp.370-397, 2009.
[22]M. Bezener and J. Hughes, Bayesian Spatiotemporal Modeling using Hierarchical Spatial Priors with Applications to Functional Magnetic Resonance Imaging, Ph.D. Dissertation, University of Minnesota Digital Conservancy, Minnesota, U.S. 2015
[23]A.A. Bakar, Z. Kefli, S. Abdullah, and M. Sahani, “Predictive models for dengue outbreak using multiple rule base classifiers,” in Proc. of Int. Conf. on Electrical Engineering and Informatics (ICEEI), Bandung, Indonesia, pp.1-6, July 2011.
[24]S. Chainey and J. Ratcliffe, GIS and Crime Map, Wiley, 2005.
[25]O. Shigueo and O. Junior, Spatiotemporal Prediction Methods, Ph.D. Dissertation, Universidad do Porto, Porto, Portugal, 2012.
[26]Sumarmo, “Dengue haemorrhagic in Indonesia,” The Southeast Asian Journal of Tropical Medicine and Public Health, Bangkok, Thailand, pp.269-274, August 1987.
連結至畢業學校之論文網頁點我開啟連結
註: 此連結為研究生畢業學校所提供,不一定有電子全文可供下載,若連結有誤,請點選上方之〝勘誤回報〞功能,我們會盡快修正,謝謝!
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top
無相關論文
 
無相關期刊
 
無相關點閱論文
 
系統版面圖檔 系統版面圖檔