請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/72545
完整後設資料紀錄
DC 欄位 | 值 | 語言 |
---|---|---|
dc.contributor.advisor | 胡明哲 | |
dc.contributor.author | Tai-Yi Liu | en |
dc.contributor.author | 劉泰億 | zh_TW |
dc.date.accessioned | 2021-06-17T07:00:39Z | - |
dc.date.available | 2022-08-07 | |
dc.date.copyright | 2019-08-07 | |
dc.date.issued | 2019 | |
dc.date.submitted | 2019-08-01 | |
dc.identifier.citation | [1] Gautam, Mr, Watanabe, K, and Saegusa, H , Journal of Hydrology, Runoff analysis in humid forest catchment with artificial neural network. 2000. 235(1-2): p. 117-136.
[2] White, Ian and Falkland, Tony, Hydrogeology Journal, Management of freshwater lenses on small Pacific islands. 2010. 18(1): p. 227-246. [3] Barnett, Jon, Regional Environmental Change, Dangerous climate change in the Pacific Islands: food production and food security. 2011. 11(1): p. 229-237. [4] Manton, Michael J, et al., Trends in extreme daily rainfall and temperature in Southeast Asia and the South Pacific: 1961–1998. 2001. 21(3): p. 269-284. [5] Grasso, Marco, Moneo, Marta, and Arena, Marco, Assessing social vulnerability to climate change in Samoa. Regional Environmental Change, 2013. 14(4): p. 1329-1341. [6] Jolliffe, Ian T, Weather, Principal component analysis: a beginner's guide—I. Introduction and application. 1990. 45(10): p. 375-382. [7] Turk, Matthew and Pentland, Alex , Journal of Cognitive Neuroscience, Eigenfaces for recognition. 1991. 3(1): p. 71-86. [8] Yildiz, Dogan, et al., Detecting seasonal cycle shift on streamflow over Turkey by using multivariate statistical methods. Theoretical and Applied Climatology, 2017. 133(3-4): p. 1143-1161. [9] Rahman, Md Habibur, Matin, M. A., and Salma, Umma, Analysis of precipitation data in Bangladesh through hierarchical clustering and multidimensional scaling. Theoretical and Applied Climatology, 2017. 134(1-2): p. 689-705. [10] Chen, Yen‐Chang, Wei, Chiang, and Yeh, Hui‐Chung , Hydrological Processes: An International Journal, Rainfall network design using kriging and entropy. 2008. 22(3): p. 340-346. [11] Tseng, Hung-Wei, et al., Application of multi-site weather generators for investigating wet and dry spell lengths under climate change: a case study in southern Taiwan. 2012. 26(15): p. 4311-4326. [12] Chen, Feng-Wen, Liu, Chen-Wuing , Paddy, and Environment, Water, Estimation of the spatial rainfall distribution using inverse distance weighting (IDW) in the middle of Taiwan. 2012. 10(3): p. 209-222. [13] Tenenbaum, Joshua B, De Silva, Vin, and Langford, John C , Science, A global geometric framework for nonlinear dimensionality reduction. 2000. 290(5500): p. 2319-2323. [14] Roweis, Sam T and Saul, Lawrence K , Science, Nonlinear dimensionality reduction by locally linear embedding. 2000. 290(5500): p. 2323-2326. [15] Böttcher, S., et al., Using Isomap to differentiate between anthropogenic and natural effects on groundwater dynamics in a complex geological setting. Journal of Hydrology, 2014. 519: p. 1634-1641. [16] Hannachi, A. and Turner, A. G., Isomap nonlinear dimensionality reduction and bimodality of Asian monsoon convection. Geophysical Research Letters, 2013. 40(8): p. 1653-1658. [17] 邱國益, 流形學習之應用的概觀研究 The Survey for Applications of Manifold Learning, in 應用數學所. 2008, 逢甲大學. [18] Torgerson, Warren S, Theory and methods of scaling. 1958. [19] Kruskal, Joseph B , Psychometrika, Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis. 1964. 29(1): p. 1-27. [20] Bernstein, Basil B, Pedagogy, symbolic control, and identity: Theory, research, critique. Vol. 5. 2000: Rowman & Littlefield. [21] Dijkstra, E. W. , Numerische Mathematik, A note on two problems in connexion with graphs. 1959. 1(1): p. 269-271. [22] Verboon, Peter and Heiser, Willem J., Resistant orthogonal procrustes analysis. Journal of Classification, 1992. 9(2): p. 237-256. [23] Schönemann, Peter H. and Carroll, Robert M. , Psychometrika, Fitting one matrix to another under choice of a central dilation and a rigid motion. 1970. 35(2): p. 245-255. [24] Wang, Chang and Mahadevan, Sridhar. Manifold alignment using procrustes analysis. in Proceedings of the 25th international conference on Machine learning. 2008. ACM. [25] Group, World Bank. Climate Change Knowledge Portal For Development Practitioners and Policy Makers. Available from: https://climateknowledgeportal.worldbank.org/. | |
dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/72545 | - |
dc.description.abstract | 在處理複雜或高維度資料時,降維是一個相當重要的課題。傳統的降維方法,如主成分分析,通常為建立在歐式空間中。然而當資料具有曲面的結構時,在歐式距離看似相近的兩點,本質上可能距離非常遠,因而在降維時造成誤差。基於流形學習的非線性降維方法,如等距特徵映射或局部線性嵌入,即是針對上述問題所發展。本研究應用等距特徵映射法分析太平洋島嶼氣候資料,以瞭解其數據的結構。在等距特徵映射中,首先根據原始距離構建鄰接圖,接著利用最短距離演算法所計算出的最短路徑來近似測地線距離。最後,將測地線距離矩陣輸入到多元尺度法中以建立能夠維持高維度裡相似性的二維座標。結果顯示第一個等距特徵映射維度與緯度有關係。另外從資料類別的敏感度分析中,我們可以推斷,對於一些國家,它們可能在空間距離上較近,但是,溫度或降雨距離可能要大得多。此外,藉由本研究我們也可以觀察資料結構隨時間的變化。 | zh_TW |
dc.description.abstract | Dimensionality reduction is an important issue while dealing with complex datasets. Traditional methods, such as principal component analysis (PCA) or multidimensional scaling (MDS), are usually based on Euclidian distance. However, this may underestimate the true distance between data points while there exists a curved structure of data. In order to solve the problem above, nonlinear dimensionality reduction methods based on manifold learning, such as isometric feature mapping (Isomap) or locally linear embedding (LLE), are proposed. In this study, Isomap is applied to help us understand the structure of climate data in the Pacific Islands. In Isomap, the neighborhood graph is constructed based on the original distance. Next, geodesic distances is approximated by the shortest path calculated by using Dijkstra’s algorithm. Finally, the geodesic distance matrix is imported into multidimensional scaling (MDS) to create a two-dimension map representing their dissimilarities in a higher dimension. Results have shown that the first Isomap dimension has a relationship with latitude. The sensitive analysis of variables has shown that though some countries are spatially close, their temperature or rainfall distance are much farther. Also, we can observe the change of data structure along time through this research. | en |
dc.description.provenance | Made available in DSpace on 2021-06-17T07:00:39Z (GMT). No. of bitstreams: 1 ntu-108-R06622020-1.pdf: 2451313 bytes, checksum: 254bc9f4de157c9b988cab867fc11ed3 (MD5) Previous issue date: 2019 | en |
dc.description.tableofcontents | 口試委員會審定書 i
誌謝 ii 摘要 iii Abstract iv Table of Contents v List of Figures vii List of Tables ix Chapter 1 Introduction 1 Chapter 2 Methodology 5 2.1 Multidimensional scaling (MDS) 5 2.2 Isometric feature mapping(Isomap) 9 2.2.1 Determination of neighborhoods and adjacent map 10 2.2.2 Geodesic distance and shortest path 14 2.3 Comparison of configurations 19 2.3.1 Multiple Solutions to MDS / Isomap 19 2.3.2 Procrustes Analysis 20 Chapter 3 Materials and Procedure 22 3.1 Study Site and data description 22 3.2 Study Process 24 Chapter 4 Results and Discussion 27 4.1 Combination of Variables at Different Ratio 27 4.2 Climate data structure varies with time 34 Chapter 5 Conclusions and Recommendations 46 5.1 Conclusions 46 5.2 Recommendations 47 Reference 48 Appendix 51 | |
dc.language.iso | en | |
dc.title | 應用等距特徵映射法分析太平洋島嶼氣候資料 | zh_TW |
dc.title | Analysis of Pacific Islands climate data using Isomap approach | en |
dc.type | Thesis | |
dc.date.schoolyear | 107-2 | |
dc.description.degree | 碩士 | |
dc.contributor.oralexamcommittee | 鄭克聲,余化龍,溫在弘,郭鴻基 | |
dc.subject.keyword | 多元尺度法,非線性降維,流形學習,等距特徵映射,太平洋島嶼氣候, | zh_TW |
dc.subject.keyword | Multidimensional scaling (MDS),Nonlinear dimension reduction,Manifold learning,Isometric feature mapping (Isomap),Pacific Islands climate data, | en |
dc.relation.page | 60 | |
dc.identifier.doi | 10.6342/NTU201902324 | |
dc.rights.note | 有償授權 | |
dc.date.accepted | 2019-08-02 | |
dc.contributor.author-college | 生物資源暨農學院 | zh_TW |
dc.contributor.author-dept | 生物環境系統工程學研究所 | zh_TW |
顯示於系所單位: | 生物環境系統工程學系 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-108-1.pdf 目前未授權公開取用 | 2.39 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。