Machine learning application to seismic site classification prediction model using Horizontal-to-Vertical Spectral Ratio (HVSR) of strong-ground motions

被引:0
|
作者
Phi, Francis G. [1 ]
Cho, Bumsu [2 ]
Kim, Jungeun [2 ]
Cho, Hyungik [3 ]
Choo, Yun Wook [1 ]
Kim, Dookie [1 ]
Kim, Inhi [4 ]
机构
[1] Kongju Natl Univ, Dept Civil & Environm Engn, 1223-24 Cheonan-daero, Cheonan Si, Chungcheongnam, South Korea
[2] Kongju Natl Univ, Dept Comp Sci & Engn, 1223-24 Cheonan-daero, Cheonan Si, Chungcheongnam, South Korea
[3] Andong Natl Univ, Dept Civil Syst Engn, 1375 Gyeongdong-ro, Andong 36729, Gyeongsangbug D, South Korea
[4] Korea Adv Inst Sci & Technol, Cho Chun Shik Grad Sch Mobil, 193 Munji-ro, Daejeon, South Korea
基金
新加坡国家研究基金会;
关键词
earthquake; machine learning; seismic design; site characterization; site classification prediction; SOIL-STRUCTURE INTERACTION; STATIONS; H/V; AMPLIFICATION; TAIWAN;
D O I
10.12989/gae.2024.37.6.539
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
This study explores development of prediction model for seismic site classification through the integration of machine learning techniques with horizontal-to-vertical spectral ratio (HVSR) methodologies. To improve model accuracy, the research employs outlier detection methods and, synthetic minority over -sampling technique (SMOTE) for data balance, and evaluates using seven machine learning models using seismic data from KiK-net. Notably, light gradient boosting method (LGBM), gradient boosting, and decision tree models exhibit improved performance when coupled with SMOTE, while Multiple linear regression (MLR) and Support vector machine (SVM) models show reduced efficacy. Outlier detection techniques significantly enhance accuracy, particularly for LGBM, gradient boosting, and voting boosting. The ensemble of LGBM with the isolation forest and SMOTE achieves the highest accuracy of 0.91, with LGBM and local outlier factor yielding the highest F1-score of 0.79. Consistently outperforming other models, LGBM proves most efficient for seismic site classification when supported by appropriate preprocessing procedures. These findings show the significance of outlier detection and data balancing for precise seismic soil classification prediction, offering insights and highlighting the potential of machine learning in optimizing site classification accuracy.
引用
收藏
页码:539 / 554
页数:16
相关论文
共 50 条
  • [41] Antarctic ice sheet thickness estimation using the horizontal-to-vertical spectral ratio method with single-station seismic ambient noise
    Yan, Peng
    Li, Zhiwei
    Li, Fei
    Yang, Yuande
    Hao, Weifeng
    Bao, Feng
    CRYOSPHERE, 2018, 12 (02): : 795 - 810
  • [42] Buried bedrock valleys revealed in Michigan's central Upper Peninsula using the horizontal-to-vertical spectral ratio passive seismic method
    VanderMeer, Sarah M.
    Bouali, El Hachemi
    Kehew, Alan E.
    Sauck, William A.
    Gillespie, Robb
    JOURNAL OF GREAT LAKES RESEARCH, 2022, 48 (05) : 1140 - 1146
  • [43] Horizontal to vertical spectral amplitude ratio of seismic waves as an effective tool for site classification: A study from Chennai, Tamilnadu
    Prasad, P. Prabhakar
    Kishore, S. Trupti P. Pavan
    Srinivas, K. N. S. S. S.
    Seshunarayana, T.
    JOURNAL OF INDIAN GEOPHYSICAL UNION, 2014, 18 (03): : 387 - 393
  • [44] Estimating the site effects in Luoyang basin using horizontal-to-vertical spectral ratio method from a short-period dense array
    Yujuan Tan
    Yunhao Wei
    Yonghong Duan
    Fuyun Wang
    Earthquake Science, 2018, (Z1) : 272 - 280
  • [45] Estimating the site effects in Luoyang basin using horizontal-to-vertical spectral ratio method from a short-period dense array
    Tan, Yujuan
    Wei, Yunhao
    Duan, Yonghong
    Wang, Fuyun
    EARTHQUAKE SCIENCE, 2018, 31 (5-6) : 272 - 280
  • [46] Soft soil mapping using Horizontal to Vertical Spectral Ratio (HVSR) for seismic hazard assessment of Chandigarh city in Himalayan foothills, north India
    A. K. Mundepi
    Conrad Lindholm
    Journal of the Geological Society of India, 2009, 74 : 551 - 558
  • [47] Soft soil mapping using Horizontal to Vertical Spectral Ratio (HVSR) for seismic hazard assessment of Chandigarh city in Himalayan foothills, north India
    Mundepi, A. K.
    Lindholm, Conrad
    Kamal
    JOURNAL OF THE GEOLOGICAL SOCIETY OF INDIA, 2009, 74 (05) : 551 - 558
  • [48] Application of horizontal-to-vertical (H/V) Fourier spectral ratio for analysis of site effect on rock (NEHRP-class B) sites in Taiwan
    Sokolov, Vladimir Yu.
    Loh, Chin-Hsiung
    Jean, Wen-Yu
    SOIL DYNAMICS AND EARTHQUAKE ENGINEERING, 2007, 27 (04) : 314 - 323
  • [49] Site response evolution and sediment mapping using horizontal to vertical spectral ratios (HVSR) of ground ambient noise in Jammu city, NW India
    Mundepi, A. K.
    Mahajan, A. K.
    JOURNAL OF THE GEOLOGICAL SOCIETY OF INDIA, 2010, 75 (06) : 799 - 806
  • [50] Site response evolution and sediment mapping using horizontal to vertical spectral ratios (HVSR) of ground ambient noise in Jammu city, NW India
    A. K. Mundepi
    A. K. Mahajan
    Journal of the Geological Society of India, 2010, 75 : 799 - 806