(3.238.173.209) 您好!臺灣時間:2021/05/15 17:22
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:::

: 
twitterline
研究生:許銘修
研究生(外文):Ming-Xiu Xu
論文名稱:基於混合粒子群優與引力搜尋演算的多數決非純度優先權法應用於高維度資料特徵抽取
論文名稱(外文):Hybrid Particle Swarm Optimization and Gravitational Search Algorithm-based Impurity Function Band Prioritization Using Weighted Majority Voting for Band Selection of High Dimensional Data Sets
指導教授:方志鵬方志鵬引用關係張陽郎張陽郎引用關係
口試委員:張陽郎方志鵬吳孟哲鞠志遠王怡鈞
口試日期:2016-07-01
學位類別:碩士
校院名稱:國立臺北科技大學
系所名稱:電機工程研究所
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
語文別:中文
中文關鍵詞:高維度資料、波段優先權、波段選取、相關係數矩陣、混合粒子群優與引力搜尋演算法。
外文關鍵詞:High Dimensional Data SetsBand PrioritizationBand SelectionCorrelation Coefficient MatrixHybrid Particle Swarm Optimization and Gravitational Search Algorithm(PSOGSA).
相關次數:
  • 被引用被引用:0
  • 點閱點閱:180
  • 評分評分:
  • 下載下載:12
  • 收藏至我的研究室書目清單書目收藏:0
隨著衛星遙測技術進步,高光譜影像(Hyperspectral Imaging)的波段數與資料量日趨龐大,在維度越來越高下計算量也隨之提高,可能使資料受到維度詛咒(Curse of Dimensionality)的影響,進而影響分類正確率,因此,需藉由波段選取來降低資料複雜度並抽取具代表性的波段。
過去已有學者提出以粒子群優(Particle Swarm Optimization, PSO)演算法應用於高光譜波段抽取,其參數設定影響粒子全域和區域搜尋甚大,在演算早期因無法搜尋較大的解空間,將使得降維效過不彰。而引力搜尋(Gravitational Search Algorithm, GSA)演算法在前期可以探索較大的解空間,但在後期因為粒子移動速度趨緩,導致無法收斂至最佳解。因此,本論文提出混合「粒子群優」的探索(Exploration) 能力和「引力搜尋」的開發(Exploitation)能力應用於高維度資料抽取,使得演算法在早期藉由引力搜尋的優點探索大範圍的解空間,後期保有粒子群優的全域最佳記憶特性而收斂至最佳解,以達高維度資料降為效果。
本論文採用Washington DC mall的HYDICE遙測影像及Northwest Tippecanoe County的AVIRIS遙測影像維實驗圖資。最後由實驗結果可以得知,本文所提出的混合粒子群優與引力搜尋演算法能夠有效的挑選出具代表性的波段降低資料為度,並透過分類器獲得良好的分類結果。
With the progress in remote sensing technologies, the number of bands and data are increased in hyperspectral imaging. The calculation become much higher in high dimensional data. The data could be facing the curse of dimensionality, that cause classification accuracy decreased. Preventing the curse of dimensionality need using band selection to reduce data computational complexity and extract the important data.
Some scholars proposed Particle Swarm Optimization algorithms for band selection to reduce dimensionality in hyperspectral imaging. The particle local and global search are affected by parameter selection. In the early stage can not exploitation well that cause the dimensionality reduction effects could not be significantly. The Gravitational Search Algorithm have good ability of exploration, but not good at exploitation. Because in the later stage particles are moved slowly that can not converge. Therefore, using hybrid particle swarm optimization and gravitational search algorithm for band selection of high dimensional data sets. PSOGSA have the PSO ability of exploitation and GSA ability of exploration to select representative bands for obtaining greatly dimensionality reduction effect.
The effectiveness of the proposed method is evaluated by HYDICE and AVIRIS remote sensing images for testing. The experimental results can be learned, “Hybrid Particle Swarm Optimization and Gravitational Search Algorithm” could reduction the dimension of data sets by band selection and have a good classification results through classification.
摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
表目錄 vi
圖目錄 vii
第一章 緒論 1
1.1 研究背景介紹 1
1.2 研究目的 2
1.3 論文內容大綱 4
第二章 相關文獻回顧 5
2.1 高光譜影像介紹 5
2.2 相關係數與相關係數矩陣 6
2.3貪婪模組特徵空間 8
2.4 粒子群優演算法 9
2.4.1 粒子群優法理論簡介 9
2.4.2 粒子群優法流程 10
2.4.3 粒子移動公式及參數 12
2.5 引力搜尋演算法 13
2.5.1 引力搜尋法理論簡介 13
2.5.2 引力搜尋法流程 14
2.5.3 引力搜尋法公式及參數 16
2.6 非純度波段優先權法 18
2.7 多數決波段選取法 19
第三章 研究方法 21
3.1 高光譜影像分類流程 21
3.2 問題說明 22
3.3 利用PSOGSA聚合相關係數矩陣高相關度波段 23
3.4 混合粒子群優與引力搜尋法公式及參數 24
3.5 適應值函數及解空間轉換 25
3.6 搜尋參數設定 26
第四章 實驗結果 28
4.1 使用圖資介紹 28
4.1.1 Washington DC mall(WDC)圖像資料介紹 28
4.1.2 Northwest Tippecanoe County(NTC)圖像資料介紹 29
4.2 實驗目的與實驗環境 31
4.3 實驗結果 31
4.3.1 相關係數矩陣的相關度數值影響 32
4.3.2 PSOGSA慣性權重影響 33
4.3.3 PSOGSA C_1^和C_2^影響 35
4.3.4 世代數影響 36
4.3.5 粒子數影響 40
第五章 結論與未來發展 44
5.1 結論 44
5.2 未來展望 44
參考文獻 45
[1]福衛二號影像精選:臺灣-和平島,國研院-國家太空中心, http://www.nspo.narl.org.tw/2011/tw/gallery/image.shtml
Accessed 5 August 2013.
[2]G. F. Hughes, "On the mean accuracy of statistical pattern recognizers," IEEE Transactions on Information Theory, vol. 14, no. 1, 1968, pp. 55-63.
[3]K. Pearson, "On Lines and Planes of Closest Fit to Systems of Points in Space," Philosophical Magazine, vol. 2, no. 6, 1901, pp. 559-572.
[4]Yang-Lang Chang, Chin-Chuan Han, Kuo-Chin Fan, K.S. Chen, Chia-Tang Chen and Jeng-Horng Chang, "Greedy Modular Eigenspaces and Positive Boolean Function for Supervised Hyperspectral Image Classification," Optical Engineering, vol. 42, no. 9, 2003, pp. 2576-2587.
[5]Yang-Lang Chang, Chin-Chuan Han, Bormin Huang, Wen-Yen Chang, Benediktsson and J.A.,Lena Chang, "A Parallel Simulated Annealing Approach to Band Selection for High-Dimensional Remote Sensing Images," IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, vol. 4, no. 3, 2011, pp. 579-590 .
[6]王旭,基於GPU之粒子群優法應用於高光譜影像波段選取,碩士論文,國立國立臺北科技大學電機工程研究所,臺北,2012。
[7]吳禎賢,高光譜影像細菌覓食優化多屬性決策波段選取使用自適應最鄰近特徵空間法分類,碩士論文,國立國立臺北科技大學電機工程研究所,臺北,2014。
[8]25 Examples of Infrared Photography, http://photography.tutsplus.com/articles/25-examples-of-infrared-photography--photo-714, Accessed 4 Nov 2009.
[9]Introduction: What is hyperspectral imagery?,
http://nyokoya.sakura.ne.jp/wordpress/research/introduction/, A ccessed 30 May 2014.
[10]C. Lee and D. A. Landgrebe, "Analyzing high-dimensional multispectral data," IEEE Transactions on Geoscience and Remote Sensing, vol. 31, no. 4, 1993, pp. 792-800.
[11]林惠玲和陳正倉,統計學-方法與應用(上冊),臺北:雙葉書廊有限公司,2000,第314-325頁。
[12]E. Rashedi, H. Nezamabadi and S. Saryazdi, "GSA: A Gravitational Search Algorithm, " Information Sciences, vol. 179, no. 13, 2009, pp. 2232-2248.
[13]Yang-Lang Chang, Chin-Chuan Han, Kuo-Chin Fan, K.S. Chen, Chia-Tang Chen and Jeng-Horng Chang, "Greedy Modular Eigenspaces and Positive Boolean Function for Supervised Hyperspectral Image Classification," Optical Engineering, vol. 42, no. 9, 2003, pp. 2576-2587.
[14]James Kennedy and Russell Eberhart, "Particle Swarm Optimization," IEEE Interantional Conference on Neural Networks, vol. 4, 1995, pp. 1942-1948.
[15]林國凱,基於粒子群優演算法的多屬性決策-非純度波段優先權方法應用於高維度資料波段選取,碩士論文,國立國立臺北科技大學電機工程研究所,臺北,2014。
[16]James Kennedy and Russell Eberhart, "Particle Swarm Optimization," IEEE Interantional Conference on Neural Networks, vol. 4, 1995, pp. 1942-1948.
[17]Yang-Lang Chang, Bin-Feng Shu, Tung-Ju Hsieh, Chih-Yuan Chu and Jyh-Perng Fang, "Band selection for hyperspectral images based on impurity function," IEEE International Geoscience and Remote Sensing Symposium (IGARSS), Vancouver, USA, 2011, pp. 2369 - 2372.
[18]Yang-Lang Chang, A Novel Approach to Hyperspectral Image Classification, Ph.D. Thesis, National Central University, Taiwan, 2003.
[19]Kenneth A De Jong, An analysis of the behavior of a class of genetic adaptive systems, Ph.D. Thesis, University of Michigan, Michigan, 1975.
[20]王宏原,平行粒子群優法應用於高維度影像特徵抽取,碩士論文,國立國立臺北科技大學電機工程研究所,臺北,2010。
[21]Leo Breiman, Jerome H. Friedman, Richard A. Olshen and Charles J. Stone, Classification and Regression Trees, USA: CRC Press, 1984, pp. 23-107.
[22]C.-I. Chang, Q. Du, T.-L. Sun, and M. L. G. Althouse, “A Joint Band Prioritization and Band Decorrelation Approach to Band Selection for Hyperspectral Image Classification,” IEEE Transactions on Geoscience and Remote Sensing, vol. 37, no. 6, 1999, pp. 2631-2641.
[23]S. Mirjalili and S. Z. M. Hashim, "A New Hybrid PSOGSA Algorithm for Function Optimization," 2010 International Conference on Computer and Information Application (ICCIA), 2010, pp. 374-377.
[24]王品皓,基於粒子群優演算法的多數決非純度波段優先權方法應用於高維度資料特徵抽取,碩士論文,國立國立臺北科技大學電機工程研究所,臺北,2013。
[25]X. Lai and M. Zhang, "An Efficient Ensemble of GA and PSO for Real Function Optimization," IEEE International Conference on Computer Science and Information Technology, 2009, pp. 651-655.
[26]A. A. A. Esmin, G. Lambert-Torres and G. B. Alvarenga, "Hybrid Evolutionary Algorithm Based on PSO and GA Mutation," Proceeding of the Sixth International Conference on Hybrid Intelligent Systems (HIS 06), 2006, pp. 57.
[27]L. Li, B. Xue, B. Niu, L. Tan, and J. Wang, "A Novel PSO-DE-Based Hybrid Algorithm for Global Optimization," Lecture Notes in Computer Science.: Springer Berlin / Heidelberg, 2008, pp. 156-163.
[28]N. Holden and A. A. Freitas, "A Hybrid PSO/ACO Algorithm for Discoverin Classification Rules in Data Mining, " Journal of Artificial Evolution and Applications (JAEA), vol. 2008, 2008.
[29]WeiPang, Kang-Ping Wang, Chun-Guang Zhou, "Modified Particle Swarm Optimization Based On Space Transformation For Solving Traveling Salesman Problem," 2004, pp. 2342-2346.
[30]R.C. Eberhart and Y. Shi., "Tracking and optimizing dynamic systems with particle swarms, "In Evolutionary Computation, 2001. Proceedings of the 2001 Congress on, vol. 1, 2002, pages 94–100.
[31]R.F. Malik, T.A. Rahman, S.Z.M. Hashim, and R. Ngah, "New Particle Swarm Optimizer with Sigmoid Increasing Inertia Weight, " International Journal of Computer Science and Security, vol. 1, no. 2, 2007, pp.35-44.
[32]J. Xin, G. Chen, and Y. Hai., "A Particle Swarm Optimizer with Multistage Linearly-Decreasing Inertia Weight, " In Computational Sciences and Optimization, 2009. CSO 2009. International Joint Conference on, vol. 1, 2009, pp 505-508.
[33]W. Al-Hassan, MB Fayek, and SI Shaheen, "PSOSA: An optimized particle swarm technique for solving the urban planning problem," International Conference on Computer Engineering and Systems,2007, pp. 401–405.
[34]Hyperspectral Image: https://engineering.purdue.edu/~biehl/MultiSpec/hyperspectral.html.
[35]Hyperspectral Remote Sensing Scenes: http://www.ehu.es/ccwintco/index.php/Hyperspectral_Remote_Sensing_Scenes.
[36]k-nearest neighbor algorithm ,Wikipedia , http://en.wikipedia.org/wiki/K-nearest_neighbor_algorithm, Accessed 19 May 2013.
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