DOI QR코드

DOI QR Code

불균형 데이터 환경에서 로지스틱 회귀모형을 이용한 Cochlodinium polykrikoides 적조 탐지 기법 연구

Study on Detection Technique for Cochlodinium polykrikoides Red tide using Logistic Regression Model under Imbalanced Data

  • 박수호 (부경대학교 지구환경시스템과학부) ;
  • 김흥민 (부경대학교 지구환경시스템과학부) ;
  • 김범규 (부경대학교 지구환경시스템과학부) ;
  • 황도현 (부경대학교 지구환경시스템과학부) ;
  • 엥흐자리갈 운자야 (부경대학교 지구환경시스템과학부) ;
  • 윤홍주 (부경대학교 지구환경시스템과학부)
  • 투고 : 2018.11.09
  • 심사 : 2018.12.15
  • 발행 : 2018.12.31

초록

본 연구에서는 불균형 데이터 환경에서 기계학습 기법의 한 갈래인 로지스틱 회귀모형을 이용하여 인공위성 영상에서 Cochlodinium polykrikoides 적조 픽셀을 탐지하는 방법을 제안한다. 학습자료로 적조, 청수, 탁수 해역에서 추출된 수출광량 분광 프로파일을 활용하였다. 전체 데이터셋의 70%를 추출하여 모형 학습에 활용하였으며, 나머지 30%를 이용하여 모형의 분류 정확도를 평가하였다. 이 때, 청수와 탁수에 비해 자료 수가 상대적으로 적은 적조의 분광 프로파일에 백색 잡음을 추가하여 오버샘플링을 하여 불균형 데이터 문제를 해결하였다. 정확도 평가 결과 본 연구에서 제안하는 알고리즘은 약 94%의 분류 정확도를 보였다.

This study proposed a method to detect Cochlodinium polykrikoides red tide pixels in satellite images using a logistic regression model of machine learning technique under Imbalanced data. The spectral profiles extracted from red tide, clear water, and turbid water were used as training dataset. 70% of the entire data set was extracted and used for as model training, and the classification accuracy of the model was evaluated using the remaining 30%. At this time, the white noise was added to the spectral profile of the red tide, which has a relatively small number of data compared to the clear water and the turbid water, and over-sampling was performed to solve the unbalanced data problem. As a result of the accuracy evaluation, the proposed algorithm showed about 94% classification accuracy.

키워드

KCTSAD_2018_v13n6_1353_f0001.png 이미지

그림 1. 연구 흐름도 Fig. 1 Flow chart of research

KCTSAD_2018_v13n6_1353_f0002.png 이미지

그림 2. 로지스틱 회귀모형을 활용한 적조탐지 과정 Fig. 2 Process of Red tide detection using Logistic regression model

KCTSAD_2018_v13n6_1353_f0003.png 이미지

그림 3. 적조 미발생 기간에 적조로 탐지된 적조 픽셀 수의 상대적 차이(A : Bak et al., 2018, B : 제안하는 알고리즘). Fig. 3 The relative difference in the number of red tide pixels detected as red tides during the no red tide occurrence period(A : Bak et al., 2018, B : Proposed Algorithm).

표 1. GOCI 밴드 구성 Table 1. Band Composition of GOCI

KCTSAD_2018_v13n6_1353_t0001.png 이미지

표 2. 혼동행렬 Table 2. Confusion Matrix

KCTSAD_2018_v13n6_1353_t0002.png 이미지

표 3. 언더샘플링을 적용한 로지스틱 회귀모형[21]의 혼동행렬 Table 3. Confusion Matrix of Logistic Regression Model applied under-sampling[21]

KCTSAD_2018_v13n6_1353_t0003.png 이미지

표 4. 제안하는 알고리즘의 혼동행렬 Table 4. Confusion Matrix of Proposed algorithm

KCTSAD_2018_v13n6_1353_t0004.png 이미지

참고문헌

  1. D. Anderson, P. Anderson, V. Bricelj, J. Cullen, and J. Rensel, Monitoring and Management Strategies for Harmful Algal Blooms in Coastal Waters. Paris: Intergovernmental Oceanographic Commission Technical Series, 2001.
  2. H. Kim, Harmful Algal Blooms in the Sea. Busan: Dasom, 2005.
  3. Y. Yoon, Sea rebellion, Red tide. Paju: Jipmoondang, 2012.
  4. Y. Kim, Y. Byun, Y. Huh, and Y. Yu, "Detection of Cochlodinium polykrikoides Red Tide Using MODIS Level 2 Data in Coastal Waters," Korean Society of Civil Engineers J. of Civil Engineering, vol. 27, no. 4D, 2007, pp. 535-540.
  5. Y. Ahn, J. Moon, W. Seo, and H. Yoon, "Inherent Optical Properties of Red Tide Algal for Ocean Color Remote Sensing Application," J. of the Korean Society for Marine Environmental Engineering, vol. 12, no. 1, 2009, pp. 47-54.
  6. S. Bak, H. Kim, B. Kim, D. Hwang, E. Unuzaya and H. Yoon, "Study on Detection Technique for Cochlodinium polykrikoides Red tide using Logistic Regression Model and Decision Tree Model," J. of the Korean institute of Electronic Communication Sciences, vol. 13, no. 4, 2018, pp. 777-786.
  7. R. Stumpf, M. Culver, P. Tester, M. Tomlinson, G. Kirkpatrick, B. Pederson, E. Truby, V. Ransibrahmanakul, and M. Soracco, "Monitoring Karenia brevis blooms in the Gulf of Mexico using satellite ocean color imagery and other data," Harmful Algae, vol. 2, no. 2, 2003, pp. 147-160. https://doi.org/10.1016/S1568-9883(02)00083-5
  8. M. Tomlinson, R. Stumpf, V. Ransibrahmanakul, E. Truby, G. Kirkpatrick, B. Pederson, G. Vargo, and C. Heil, "Evaluation of the use of SeaWiFS imagery for detecting Karenia brevis harmful algal blooms in the eastern Gulf of Mexico," Remote Sensing of Environment, vol. 91, no. 3, 2004, pp. 293-303. https://doi.org/10.1016/j.rse.2004.02.014
  9. Y. Suh, L. Jang, N. Lee, and J. Ishizaka, "Feasibility of Red Tide Detection Around Korean Waters Using Satellite Remote Sensing," J. of Fisheries Science and Technology, vol. 7, no. 3, 2004, pp. 148-162.
  10. J. Ishizaka, Y. Kitaura, Y. Touke, H. Sasaki, A. Tanaka, H. Murakami, T. Suzuki, K. Matsuoka, and H. Nakata, "Satellite Detection of Red Tide in Ariake Sound, 1998-2001," J. of Oceanography, vol. 62, no. 1, 2006, pp. 37-45. https://doi.org/10.1007/s10872-006-0030-1
  11. Y. Son, Y. Kang, and J. Ryu, "Monitoring Red Tide in South Sea of Korea(SSK) Using the Geostationary Ocean Color Imager(GOCI)," Korean J. of Remote Sensing, vol. 26, no. 5, 2012, pp. 531-548.
  12. Y. Ahn and P. Shanmugam, "Detecting the red tide algal bloom from satellite ocean color observations in optically complex Northeast-Asia Coastal waters," Remote Sensing of Environment, vol. 103, no. 4, 2006, pp. 419-437. https://doi.org/10.1016/j.rse.2006.04.007
  13. Y. Son, J. Ishizaka, J. Jeong, H. Kim, and T. Lee, "Cochlodinium polykrikoides red tide detection in the South Sea of Korea using spectral classification of MODIS data," Ocean Science J., vol. 46, no. 4, 2011, pp. 239-263. https://doi.org/10.1007/s12601-011-0019-6
  14. S. Bak, H. Kim, D. H. Hwang, H. Yoon, and W. Seo, "Detection technique of Red tide Using GOCI Level 2 Data," Korean J. Remote Sensing, vol. 32, no. 6, 2016, pp. 673-679. https://doi.org/10.7780/KJRS.2016.32.6.11
  15. S. Bak, H. Kim, D. Hwang, S. Oh, and H. Yoon, "Red Tide Detection Technique by Using Multi-temporal GOCI Level 2 Data," Int. J. of Grid and Distributed Computing, vol. 10, no. 10, 2017, pp. 45-56. https://doi.org/10.14257/ijgdc.2017.10.10.04
  16. S. Bak and H. Yoon, "Analysis on optical property in the South Sea of Korea by using Satellite Image : Study of Case on red tide occurrence in August 2013," J. of the Korean institute of Electronic Communication Sciences, vol. 11, no. 7, 2016, pp. 723-728. https://doi.org/10.13067/JKIECS.2016.11.7.723
  17. H. Kim, S. Jang, and H. Yoon, "Utilization of Unmanned Aerial Vehicle(UAV) Image for Detection of Algal Bloom in Nakdong River," J. of the Korean institute of Electronic Communication Sciences, vol. 12, no. 3, 2017, pp. 457-464.
  18. B. Chae, W. Kim, Y. Cho, K. Kim, C. Lee, and Y. Choi, "Development of a Logistic Regression Model for Probabilistic Prediction of Debris Flow," The J. of Engineering Geology, vol. 14, no. 2, 2014, pp. 211-222.
  19. C. Park, Y. Kim, J. Kim, J. Song, and H. Choi, Data-mining using R. Seoul: Kyowoo, 2013.
  20. S. Oh, J. Park, and H. Yoon, "Prediction of Red Tide Occurrence by using Oceanic and Atmospheric Data by Satellite," J. of the Korean institute of Electronic Communication Sciences, vol. 10, no. 2, 2015, pp. 311-318. https://doi.org/10.13067/JKIECS.2015.10.2.311
  21. S. Bak, H. Kim, B. Kim, D. Hwang, U. Enkhjargal, and H. Yoon, "Study on Detection Technique for Cochlodinium polykrikoides Red tide using Logistic Regression Model and Decision Tree Model," J. of the Korean institute of Electronic Communication Sciences, vol. 13, no. 4, 2018, pp. 777-786.
  22. S. Bak, D. Hwang, H. Kim, B. Kim, U. Enkgjargal, S. Oh, and H. Yoon, "A Study on Red Tide Detection Technique by using Multi-Layer Perceptron," Int. J. of Grid and Distributed Computing, vol. 11, no. 9, 2018, pp. 93-102.
  23. P. Kang, H. Lee and S. Cho, "SVM Ensemble Techniques for Class Imbalance Problem," Proc. of Korea Information Science Society Conference, vol. 31, no. 2, Korea, 2004, pp. 706-708.