R. Adnan, Zhongmin Liang, Alban Kuriqi, O. Kisi, Anurag Malik, Binquan Li
{"title":"使用启发式机器学习方法进行流量预测","authors":"R. Adnan, Zhongmin Liang, Alban Kuriqi, O. Kisi, Anurag Malik, Binquan Li","doi":"10.1109/ICCIS49240.2020.9257658","DOIUrl":null,"url":null,"abstract":"Streamflow forecasting is vital for designing and managing water resources systems. This study evaluates the prediction accuracy of two heuristic methods, artificial neural network-genetic algorithm (ANN-GA) and adaptive neurofuzzy inference system-genetic algorithm (ANFIS-GA) in streamflow prediction using monthly streamflow data of Neelum and Kunhar Rivers of Pakistan. The prediction capability of two methods are tested using the different time lags input combinations using statistical indicators and compared with M5 Regression Tree (M5RT) model. In results, it is found that ANN-GA and ANFIS-GA provided better prediction accuracy than M5RT model. Addition of month number showed a positive effect of periodicity on the prediction accuracy of models.","PeriodicalId":425637,"journal":{"name":"2020 2nd International Conference on Computer and Information Sciences (ICCIS)","volume":"151 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"2","resultStr":"{\"title\":\"Streamflow forecasting using heuristic machine learning methods\",\"authors\":\"R. Adnan, Zhongmin Liang, Alban Kuriqi, O. Kisi, Anurag Malik, Binquan Li\",\"doi\":\"10.1109/ICCIS49240.2020.9257658\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Streamflow forecasting is vital for designing and managing water resources systems. This study evaluates the prediction accuracy of two heuristic methods, artificial neural network-genetic algorithm (ANN-GA) and adaptive neurofuzzy inference system-genetic algorithm (ANFIS-GA) in streamflow prediction using monthly streamflow data of Neelum and Kunhar Rivers of Pakistan. The prediction capability of two methods are tested using the different time lags input combinations using statistical indicators and compared with M5 Regression Tree (M5RT) model. In results, it is found that ANN-GA and ANFIS-GA provided better prediction accuracy than M5RT model. Addition of month number showed a positive effect of periodicity on the prediction accuracy of models.\",\"PeriodicalId\":425637,\"journal\":{\"name\":\"2020 2nd International Conference on Computer and Information Sciences (ICCIS)\",\"volume\":\"151 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-10-13\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"2\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2020 2nd International Conference on Computer and Information Sciences (ICCIS)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/ICCIS49240.2020.9257658\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 2nd International Conference on Computer and Information Sciences (ICCIS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICCIS49240.2020.9257658","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Streamflow forecasting using heuristic machine learning methods
Streamflow forecasting is vital for designing and managing water resources systems. This study evaluates the prediction accuracy of two heuristic methods, artificial neural network-genetic algorithm (ANN-GA) and adaptive neurofuzzy inference system-genetic algorithm (ANFIS-GA) in streamflow prediction using monthly streamflow data of Neelum and Kunhar Rivers of Pakistan. The prediction capability of two methods are tested using the different time lags input combinations using statistical indicators and compared with M5 Regression Tree (M5RT) model. In results, it is found that ANN-GA and ANFIS-GA provided better prediction accuracy than M5RT model. Addition of month number showed a positive effect of periodicity on the prediction accuracy of models.