Application of machine learning to an early warning system for very short-term heavy rainfall

被引:62
作者
Moon, Seung-Hyun [1 ]
Kim, Yong-Hyuk [2 ]
Lee, Yong Hee [3 ]
Moon, Byung-Ro [1 ]
机构
[1] Seoul Natl Univ, Sch Comp Sci & Engn, 1 Gwanak Ro, Seoul 08826, South Korea
[2] Kwangwoon Univ, Sch Software, 20 Kwangwoon Ro, Seoul 01897, South Korea
[3] Korea Meteorol Adm, Numer Modeling Ctr, Numer Data Applicat Div, 61 Yeouidaebang Ro 16 Gil, Seoul 07062, South Korea
基金
新加坡国家研究基金会;
关键词
Early warning system; Heavy rainfall nowcasting; Machine learning; Discretization; Classification; NEURAL-NETWORKS; FLOOD; DISCRETIZATION; PRECIPITATION; INFORMATION; PREDICTION; SELECTION; MODEL;
D O I
10.1016/j.jhydrol.2018.11.060
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
The purpose of an early warning system (EWS) is to issue warning signals prior to extreme events. Extreme weather events, however, are hard to predict due to their chaotic behavior. This paper suggests a method for an effective EWS for very short-term heavy rainfall with machine learning techniques. The EWS produces a warning signal when it is expected to reach the criterion for a heavy rain advisory within the next 3 h. We devised a selective discretization method that converts a subset of continuous input variables to nominal ones. Meteorological data obtained from automatic weather stations are preprocessed by the selective discretization and principal component analysis. As a classifier, logistic regression is used to predict whether or not a warning is required. A comparative evaluation was performed on the EWS models generated by various classifiers. The tests were run for 652 locations in South Korea from 2007 to 2012. The empirical results showed that the preprocessing methods improved the prediction quality and logistic regression works well on heavy rainfall nowcasting in terms of F-measure and equitable threat score.
引用
收藏
页码:1042 / 1054
页数:13
相关论文
共 72 条
[41]  
Lee J, 2012, JOINT INT CONF SOFT, P830, DOI 10.1109/SCIS-ISIS.2012.6505383
[42]   Detecting Anomalies in Meteorological Data Using Support Vector Regression [J].
Lee, Min-Ki ;
Moon, Seung-Hyun ;
Yoon, Yourim ;
Kim, Yong-Hyuk ;
Moon, Byung-Ro .
ADVANCES IN METEOROLOGY, 2018, 2018
[43]   Feature Selection: A Data Perspective [J].
Li, Jundong ;
Cheng, Kewei ;
Wang, Suhang ;
Morstatter, Fred ;
Trevino, Robert P. ;
Tang, Jiliang ;
Liu, Huan .
ACM COMPUTING SURVEYS, 2018, 50 (06)
[44]   Discretization: An enabling technique [J].
Liu, H ;
Hussain, F ;
Tan, CL ;
Dash, M .
DATA MINING AND KNOWLEDGE DISCOVERY, 2002, 6 (04) :393-423
[45]   An improved Naive Bayesian classifier technique coupled with a novel input solution method [J].
Liu, JNK ;
Li, BNL ;
Dillon, TS .
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART C-APPLICATIONS AND REVIEWS, 2001, 31 (02) :249-256
[46]   Exploratory Undersampling for Class-Imbalance Learning [J].
Liu, Xu-Ying ;
Wu, Jianxin ;
Zhou, Zhi-Hua .
IEEE TRANSACTIONS ON SYSTEMS MAN AND CYBERNETICS PART B-CYBERNETICS, 2009, 39 (02) :539-550
[47]   ESTABLISHING A ROBUST TECHNIQUE FOR MONITORING AND EARLY WARNING OF FOOD INSECURITY IN POST-CONFLICT SOUTH SUDAN USING ORDINAL LOGISTIC REGRESSION [J].
Lokosang, L. B. ;
Ramroop, S. ;
Hendriks, S. L. .
AGREKON, 2011, 50 (04) :101-130
[48]   Improving the nowcasting of precipitation in an Alpine region with an enhanced radar echo tracking algorithm [J].
Mecklenburg, S ;
Joss, J ;
Schmid, W .
JOURNAL OF HYDROLOGY, 2000, 239 (1-4) :46-68
[49]   Comparison of four machine learning algorithms for their applicability in satellite-based optical rainfall retrievals [J].
Meyer, Hanna ;
Kuehnlein, Meike ;
Appelhans, Tim ;
Nauss, Thomas .
ATMOSPHERIC RESEARCH, 2016, 169 :424-433
[50]  
Murphy KP, 2012, MACHINE LEARNING: A PROBABILISTIC PERSPECTIVE, P1