Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation
We developed a method to predict heavy rainfall in South Korea with a lead time of one to six hours. We modified the AWS data for the recent four years to perform efficient prediction, through normalizing them to numeric values between 0 and 1 and undersampling them by adjusting the sampling sizes o...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2014-01-01
|
| Series: | Advances in Meteorology |
| Online Access: | http://dx.doi.org/10.1155/2014/203545 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832568104342781952 |
|---|---|
| author | Jae-Hyun Seo Yong Hee Lee Yong-Hyuk Kim |
| author_facet | Jae-Hyun Seo Yong Hee Lee Yong-Hyuk Kim |
| author_sort | Jae-Hyun Seo |
| collection | DOAJ |
| description | We developed a method to predict heavy rainfall in South Korea with a lead time of one to six hours. We modified the AWS data for the recent four years to perform efficient prediction, through normalizing them to numeric values between 0 and 1 and undersampling them by adjusting the sampling sizes of no-heavy-rain to be equal to the size of heavy-rain. Evolutionary algorithms were used to select important features. Discriminant functions, such as support vector machine (SVM), k-nearest neighbors algorithm (k-NN), and variant k-NN (k-VNN), were adopted in discriminant analysis. We divided our modified AWS data into three parts: the training set, ranging from 2007 to 2008, the validation set, 2009, and the test set, 2010. The validation set was used to select an important subset from input features. The main features selected were precipitation sensing and accumulated precipitation for 24 hours. In comparative SVM tests using evolutionary algorithms, the results showed that genetic algorithm was considerably superior to differential evolution. The equitable treatment score of SVM with polynomial kernel was the highest among our experiments on average. k-VNN outperformed k-NN, but it was dominated by SVM with polynomial kernel. |
| format | Article |
| id | doaj-art-c4166f2be118494ca64511c46925481d |
| institution | Kabale University |
| issn | 1687-9309 1687-9317 |
| language | English |
| publishDate | 2014-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Advances in Meteorology |
| spelling | doaj-art-c4166f2be118494ca64511c46925481d2025-02-03T00:59:46ZengWileyAdvances in Meteorology1687-93091687-93172014-01-01201410.1155/2014/203545203545Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary ComputationJae-Hyun Seo0Yong Hee Lee1Yong-Hyuk Kim2Department of Computer Science and Engineering, Kwangwoon University, 20 Kwangwoon-Ro, Nowon-Gu, Seoul 139-701, Republic of KoreaForecast Research Laboratory, National Institute of Meteorological Research, Korea Meteorological Administration, 45 Gisangcheong-gil, Dongjak-gu, Seoul 156-720, Republic of KoreaDepartment of Computer Science and Engineering, Kwangwoon University, 20 Kwangwoon-Ro, Nowon-Gu, Seoul 139-701, Republic of KoreaWe developed a method to predict heavy rainfall in South Korea with a lead time of one to six hours. We modified the AWS data for the recent four years to perform efficient prediction, through normalizing them to numeric values between 0 and 1 and undersampling them by adjusting the sampling sizes of no-heavy-rain to be equal to the size of heavy-rain. Evolutionary algorithms were used to select important features. Discriminant functions, such as support vector machine (SVM), k-nearest neighbors algorithm (k-NN), and variant k-NN (k-VNN), were adopted in discriminant analysis. We divided our modified AWS data into three parts: the training set, ranging from 2007 to 2008, the validation set, 2009, and the test set, 2010. The validation set was used to select an important subset from input features. The main features selected were precipitation sensing and accumulated precipitation for 24 hours. In comparative SVM tests using evolutionary algorithms, the results showed that genetic algorithm was considerably superior to differential evolution. The equitable treatment score of SVM with polynomial kernel was the highest among our experiments on average. k-VNN outperformed k-NN, but it was dominated by SVM with polynomial kernel.http://dx.doi.org/10.1155/2014/203545 |
| spellingShingle | Jae-Hyun Seo Yong Hee Lee Yong-Hyuk Kim Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation Advances in Meteorology |
| title | Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation |
| title_full | Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation |
| title_fullStr | Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation |
| title_full_unstemmed | Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation |
| title_short | Feature Selection for Very Short-Term Heavy Rainfall Prediction Using Evolutionary Computation |
| title_sort | feature selection for very short term heavy rainfall prediction using evolutionary computation |
| url | http://dx.doi.org/10.1155/2014/203545 |
| work_keys_str_mv | AT jaehyunseo featureselectionforveryshorttermheavyrainfallpredictionusingevolutionarycomputation AT yongheelee featureselectionforveryshorttermheavyrainfallpredictionusingevolutionarycomputation AT yonghyukkim featureselectionforveryshorttermheavyrainfallpredictionusingevolutionarycomputation |