DOI QR코드

DOI QR Code

심층신경망을 활용한 Cochlodinium polykrikoides 적조 발생 예측 연구

Study on Cochlodinium polykrikoides Red tide Prediction using Deep Neural Network under Imbalanced Data

  • 박수호 (부경대학교 지구환경시스템과학부) ;
  • 정민지 (부경대학교 지구환경시스템과학부) ;
  • 황도현 (부경대학교 지구환경시스템과학부) ;
  • 엥흐자리갈 운자야 (부경대학교 지구환경시스템과학부) ;
  • 김나경 (부경대학교 지구환경시스템과학부) ;
  • 윤홍주 (부경대학교 지구환경시스템과학부)
  • 투고 : 2019.09.30
  • 심사 : 2019.12.15
  • 발행 : 2019.12.31

초록

본 연구에서는 심층 신경망을 이용하여 Cochlodinium polykrikoides 적조 발생을 예측하는 모델을 제안한다. 적조 발생 예측을 위해 8개의 은닉층을 가진 심층 신경망을 구축하였다. 위성 재분석 자료와 기상수치모델 자료를 이용하여 과거 적조 발생해역의 해양 및 기상인자 총 59개를 추출하여 신경망 모델 학습에 활용하였다. 전체 데이터셋 중 적조 발생 사례는 적조 미발생 사례에 비해 매우 적어 불균형 데이터 문제가 발생하였다. 본 연구에서는 이를 해결하기 위해 과표집화(Over sampling) 기반 데이터 증식(Data augmentation) 기법을 적용하였다. 과거자료를 활용하여 모형의 정확도를 평가한 결과 약 97%의 정확도를 보였다.

In this study, we propose a model for predicting Cochlodinium polykrikoides red tide occurrence using deep neural networks. A deep neural network with eight hidden layers was constructed to predict red tide occurrence. The 59 marine and meteorological factors were extracted and used for neural network model training using satellite reanalysis data and meteorological model data. The red tide occurred in the entire dataset is very small compared to the case of no red tide, resulting in an unbalanced data problem. In this study, we applied over sampling with adding noise based data augmentation to solve this problem. As a result of evaluating the accuracy of the model using test data, the accuracy was about 97%.

키워드

참고문헌

  1. H. Kim, C. Moon, and H. Cho, "Spacial-Temporal Characteristics of Dinoflagellate Cyst Distribution in Sediments of Busan Harbor," The Sea, vol. 10, no. 4, 2005, pp. 196-203.
  2. S. Kim, W. Go, Y. Jo, and K. Jeon, "Low Salinity Anomaly and Nutrient Distribution at Surface Waters of the South Sea of Korea during 1996 Summer," The Sea, vol. 3, no. 3, 1998, pp. 165-169.
  3. Y. Yoon, Sea Rebellion, Red Tide. Paju: Jipmoondang, 2012.
  4. 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, 2012, pp. 311-318.
  5. S. Park, K. Kim, J. Lee, and S. Lee, "Red Tide Prediction using Neural Network and SVM," J. of the Institute of Electronics Engineers of Korea, vol. 48SP, no. 5, 2011, pp. 651-657.
  6. C. C. Aggarwal, Neural Networks and Deep Learning : A Textbook. Basel: Springer, 2018.
  7. I. Goodfellow, Y. Bengio, and A. Courville, Deep Learning. Massachusetts: MIT Press book, 2016.
  8. J. C. M. Pires, F. G. Martins, S. I. V. Sousa, M. C. M. Alvim-Ferraz, and M. C. Pereira, "Selection and validation of parameters in multiple linear and principal component regression," Environmental Modelling & Software, vol. 23, no. 1, pp. 50-55. https://doi.org/10.1016/j.envsoft.2007.04.012
  9. C. De Mol, D. Giannone, and L. Reichlin, "Forecasting using a large number of predictors: Is Bayesian shrinkage a valid alternative to principal components?," J. of Econometrics, vol. 146, no. 2, 2008, pp. 318-328. https://doi.org/10.1016/j.jeconom.2008.08.011
  10. Y. Kim, Y. Byun, Y. Huh, and Y. Yu, "Detection of Cochlodinium polykrikoides Red Tide Using MODIS Level 2 Data in Coastal Waters," KSCE Journal of Civil Engineering, vol. 27, no. 4D, 2007, pp. 535-540.
  11. G. Cybenko, "Approximation by Superpositions of a Sigmoidal Function," Mathematics of Control, Signals and Systems, vol. 2, no. 4, 1989, pp. 303-314. https://doi.org/10.1007/BF02551274