DOI QR코드

DOI QR Code

Categorical Prediction and Improvement Plan of Snow Damage Estimation using Random Forest

랜덤포레스트를 이용한 대설피해액에 대한 범주형 예측 및 개선방안 검토

  • Received : 2019.04.22
  • Accepted : 2019.05.16
  • Published : 2019.05.31

Abstract

Recently, the occurrence of unusual heavy snow and cold are increasing due to the unusual global climate change. In particular, the temperature dropped to minus 69 degrees Celsius in the United States on January 8, 2018. In Korea, on February 17, 2014, the auditorium building in Gyeongju Mauna Resort was collapsed due to the heavy snowfall. Because of the tragic accident many studies on the reduction of snow damage is being conducted, but it is difficult to predict the exact damage due to the lack of historical damage data, and uncertainty of meteorological data due to the long distance between the damaged area and the observatory. Therefore, in this study, available data were collected from factors that are thought to be corresponding to snow damage, and the amount of snow damage was estimated categorically using a random forest. At present, the prediction accuracy was not sufficient due to lack of historical damage data and changes of the design code for green houses. However, if accurate weather data are obtained in the affected areas. the accuracy of estimates would increase enough for being used for be the degree preparedness of disaster management.

최근 세계적인 기상이변으로 이례적인 대설과 한파의 발생 빈도가 증가하고 있다. 이로 인해 대설피해 저감에 대한 연구가 진행되고 있으나, 우리나라는 시군구 별 과거 피해이력이 적고, 피해 발생지역과 관측소 간의 거리가 멀어 정확한 피해예측이 어려운 상황이다. 따라서 본 연구에서는 대설피해에 영향을 미친다고 생각되는 인자들의 데이터를 수집한 뒤 랜덤포레스트 모형의 설명변수로 설정하여 추정되는 대설피해액을 범주형태로 예측하고자 하였다. 또한 설명변수 중 취약성 분석을 통해 도출된 취약성 지수를 설명변수로 이용함으로써 지역적 특색과 특성을 반영하였다. 지금은 과거 피해 자료의 부족, 비닐하우스 설계 기준의 변화 등으로 인해 예측 정확도가 높지 않지만, 피해가 발생한 지역의 정확한 기상자료가 확보되고, 변수로 사용한 데이터의 업데이트가 진행된다면 본 연구 결과의 정확도 향상과 재난 발생 전 피해규모 및 범위에 대한 신속한 예측을 통해 대비차원의 재난관리 대처능력이 향상될 것으로 기대된다.

Keywords

HKSJBV_2019_v21n2_157_f0001.png 이미지

Fig. 1. Random Forest Structure

HKSJBV_2019_v21n2_157_f0002.png 이미지

Fig. 2. P-S-R Structure System

HKSJBV_2019_v21n2_157_f0003.png 이미지

Fig. 3. Property Information Matrix

HKSJBV_2019_v21n2_157_f0004.png 이미지

Fig. 4. Entropy Weight Estimation Process

HKSJBV_2019_v21n2_157_f0005.png 이미지

Fig. 5. Calculated Snow Vulnerability Index

HKSJBV_2019_v21n2_157_f0006.png 이미지

Fig. 6. Importance of Variables

HKSJBV_2019_v21n2_157_f0007.png 이미지

Fig. 7. Accuracy and Error Rate

Table 1. Variable Name and Description

HKSJBV_2019_v21n2_157_t0001.png 이미지

Table 2. Input Data for Snow Vulnerability Index

HKSJBV_2019_v21n2_157_t0002.png 이미지

Table 3. Historical Number of Snow Damage in terms of Damage Categories

HKSJBV_2019_v21n2_157_t0003.png 이미지

References

  1. Breiman, L, (2001). Random forests machine learning, 45(1), pp. 5-32. DOI : https://doi.org/10.1023/A:1010933404324
  2. Choi, CH, Park, KH, Park, HK, Lee, MG, Kim, JS, Kim, HS, (2017). Development of heavy rain damage prediction function for public facility using machine learning, J. of Korean Society of Hazard Mitigation, 17(6), pp. 443-450. [Korean Literature]. DOI : https://doi.org/10.9798/KOSHAM.2017.17.6.443
  3. Ha, R, Shin, HJ, Kim, SJ, (2007). Proposal of prediction technique for future vegetation information by climate change using satellite image, J. of Korean Association of Geographic Information Studies, 10(3), pp. 58-69. [Korean Literature] DOI : http://www.koreascience.or.kr/article/JAKO200721761942397.page
  4. Kwon, SH, Park, HS, Chung, GH, (2016). Analysis of snow vulnerability and adaptation policy for heavy snow, J. of Korean Society of Hazard Mitigation, 16(2), pp. 363-368. [Korean Literature]. DOI : https://doi.org/10.9798/KOSHAM.2016.16.2.363
  5. Kwon, SH, Chung, GH, (2017). Estimation of snow damages using multiple regression model: The Case of Gangwon Province, J. of the Korean Society of Civil Engineers, 37(1), pp. 61-72. [Korean Literature]. DOI: https://doi.org/10.9798/KOSHAM.2016.16.2.437
  6. MPSS(Ministry of Public Safety and Security)(2014). The 2013 Annual Natural Disaster report, Ministry of Public Safety and Security. [Korean Literature]
  7. Oh, YR, Chung, GH, (2017). Estimation of snow damage and proposal of snow damage threshold based on historical disaster data, J. of the Korean Society of Civil Engineers, 37(2), pp. 325-331. [Korean Literature]. DOI : https://doi.org/10.12652/ksce.2017.37.2.0325
  8. Oh, YR, Chung, GH, (2018). Multiple regression models of snow damage prediction according to the snow damage vulnerability groups Korea, J. of Korean Society of Hazard Mitigation, 18(2), pp. 355-359. [Korean Literature]. DOI: https://doi.org/10.9798/KOSHAM.2018.18.2.355
  9. Shannon, CE, Weaver, W, (1949). The mathematical theory of communication, The University of Illinois Press, Urbana, U.S.A