不同数据处理策略对Chl-a浓度预测精度的影响
|
摘要
为了确定合理数据处理策略,提高基于神经网络的叶绿素a含量预测精度,采用7种数据处理方案和5种神经网络输入参数组合,研究了不同数据处理策略对叶绿素a含量预测精度的影响。结果表明:数据处理可有效提高基于神经网络的水体中叶绿素a含量预测精度,不同数据处理策略得到的主成分不同,对预测精度的提高程度不同;在输入参数数量相同情况下,以格拉布斯准则处理异值点,再采用局部多项式回归进行数据平滑所得的神经网络预测精度最高;4个输入参数情况下,预测精度在进行数据处理后最高可达到0.986,比采用原始数据提高23.25%。
To determine a reasonable data processing strategy for improving the prediction accuracy of chlorophyll-a content based on neural network, the effects of different data processing strategies were studied by using seven data processing schemes and five neural network input parameters. The results show that data processing can effectively improve the prediction accuracy of chlorophyll-a content in water. Different data processing strategies have different main components, thus contributing different improving effects to the prediction. In the condition of inputting the same parameters, the neural network prediction accuracy is maximized by treating the outliers with the Grubbs criterion and using "Savitzky-Golay" to smooth the data. In the case of inputting 4 parameters, the prediction accuracy can reach 0.986 after data processing, which is 23.25% higher than directly using the original data.
To determine a reasonable data processing strategy for improving the prediction accuracy of chlorophyll-a content based on neural network, the effects of different data processing strategies were studied by using seven data processing schemes and five neural network input parameters. The results show that data processing can effectively improve the prediction accuracy of chlorophyll-a content in water. Different data processing strategies have different main components, thus contributing different improving effects to the prediction. In the condition of inputting the same parameters, the neural network prediction accuracy is maximized by treating the outliers with the Grubbs criterion and using "Savitzky-Golay" to smooth the data. In the case of inputting 4 parameters, the prediction accuracy can reach 0.986 after data processing, which is 23.25% higher than directly using the original data.
引文
[1] 卢小燕,徐福留,詹巍,等.湖泊富营养化模型的研究现状与发展趋势[J].水科学进展,2003,14(6):792-798.
[2] 裴洪平,罗妮娜,蒋勇.利用BP神经网络方法预测西湖叶绿素a的浓度[J].生态学报,2004,24(2):246-251.
[3] Rafiee Mohammad,Jahangiri-Rad Mahsa.Artificial Neural Network Approaches to the Prediction of Eutrophication and Algal Blooms in Aras Dam [J].Ulūm-i Bihdāshtī-i īrān,2015,3(1):25-32.
[4] Bui Manh-Ha,Pham Thanh-Luu,Dao Thanh-Son.Prediction of cyanobacterial blooms in the Dau Tieng Reservoir using an artificial neural network[J].Marine and Freshwater Research,2017,68(11):2070-2080.
[5] Recknagel Friedrich.ANNA -- Artificial Neural Network for predicting species abundance and succession of blue-green algae[J].Hydrobiologia,1997,349(1-3):47.
[6] Velo-Suarez L,Gutierrez-Estrada J C.Artificial neural network approaches to one-step weekly prediction of Dinophysis acuminata blooms in Huelva (Western Andalucia,Spain)[J].Harmful Algae,2007,6(3):361-371.
[7] C Guallar,M Delgado.Artificial neural network approach to population dynamics of harmful algal blooms in Alfacs Bay (NW Mediterranean):Case studies of Karlodinium and Pseudo-nitzschia[J].Ecological Modelling,2016,338(10):37-50.
[8] 赵文喜,周滨,刘红磊,等.基于BP神经网络的海河干流叶绿素浓度短时预测研究[J].水利水电技术,2017,48(11):134-140.
[9] 方洪鹰.数据挖掘中数据预处理的方法研究[D].重庆:西南大学,2009.
[10] 王亚宸.基于数据预处理和人工智能优化的组合预测模型的研究及应用[D].兰州:兰州大学,2017.
[11] Iliou Theodoros,Anagnostopoulos Christos-Nikolaos,et al.A novel data preprocessing method for boosting neural network performance:a case study in osteoporosis prediction[J].Information Sciences,2017,380(2):92-100.
[12] 李丽敏,温宗周,王真,等.基于自学习Pauta和smooth的地下水位异常值检测和平滑处理方法[J].西安工程大学学报,2018,32(5):604-608.
[13] Sanches Regina Aparecida,Takamune Karina Mitie.Comparative Study of the Characteristics of Knitted Fabrics Produced from Recycled Fibres Employing the Chauvenet Criterion,Factorial Design and Statistical Analysis[J].Fibres & Textiles in Eastern Europe,2015,23(4):19-24.
[14] 池庆玺,司锡才.基于格拉布斯准则的雷达信号分选方法的探讨[J].传感技术学报,2006,19(6):2625-2629.
[15] Acharya D,Rani A,Agarwal S,et al.Application of adaptive Savitzky-Golay filter for EEG signal processing[J].Perspectives in Science,2016,8(C):677-679.
[16] 刘翔.BP算法的改进及其应用[D].太原:太原理工大学,2012.
[17] 张颖,李彩娟,邵惠鹤.基于主成分分析与模糊BP方法的藻类繁殖状态预测[J].上海交通大学学报.2012,46(5):785-789.
[18] 段洪涛,张柏,宋开山,等.长春市南湖叶绿素a浓度高光谱估测模型研究[J].水科学进展.2006,17(3):323-329.
[19] 何沁波.龙景湖叶绿素a浓度预测模型敏感性分析[D].重庆:重庆大学,2015.
[2] 裴洪平,罗妮娜,蒋勇.利用BP神经网络方法预测西湖叶绿素a的浓度[J].生态学报,2004,24(2):246-251.
[3] Rafiee Mohammad,Jahangiri-Rad Mahsa.Artificial Neural Network Approaches to the Prediction of Eutrophication and Algal Blooms in Aras Dam [J].Ulūm-i Bihdāshtī-i īrān,2015,3(1):25-32.
[4] Bui Manh-Ha,Pham Thanh-Luu,Dao Thanh-Son.Prediction of cyanobacterial blooms in the Dau Tieng Reservoir using an artificial neural network[J].Marine and Freshwater Research,2017,68(11):2070-2080.
[5] Recknagel Friedrich.ANNA -- Artificial Neural Network for predicting species abundance and succession of blue-green algae[J].Hydrobiologia,1997,349(1-3):47.
[6] Velo-Suarez L,Gutierrez-Estrada J C.Artificial neural network approaches to one-step weekly prediction of Dinophysis acuminata blooms in Huelva (Western Andalucia,Spain)[J].Harmful Algae,2007,6(3):361-371.
[7] C Guallar,M Delgado.Artificial neural network approach to population dynamics of harmful algal blooms in Alfacs Bay (NW Mediterranean):Case studies of Karlodinium and Pseudo-nitzschia[J].Ecological Modelling,2016,338(10):37-50.
[8] 赵文喜,周滨,刘红磊,等.基于BP神经网络的海河干流叶绿素浓度短时预测研究[J].水利水电技术,2017,48(11):134-140.
[9] 方洪鹰.数据挖掘中数据预处理的方法研究[D].重庆:西南大学,2009.
[10] 王亚宸.基于数据预处理和人工智能优化的组合预测模型的研究及应用[D].兰州:兰州大学,2017.
[11] Iliou Theodoros,Anagnostopoulos Christos-Nikolaos,et al.A novel data preprocessing method for boosting neural network performance:a case study in osteoporosis prediction[J].Information Sciences,2017,380(2):92-100.
[12] 李丽敏,温宗周,王真,等.基于自学习Pauta和smooth的地下水位异常值检测和平滑处理方法[J].西安工程大学学报,2018,32(5):604-608.
[13] Sanches Regina Aparecida,Takamune Karina Mitie.Comparative Study of the Characteristics of Knitted Fabrics Produced from Recycled Fibres Employing the Chauvenet Criterion,Factorial Design and Statistical Analysis[J].Fibres & Textiles in Eastern Europe,2015,23(4):19-24.
[14] 池庆玺,司锡才.基于格拉布斯准则的雷达信号分选方法的探讨[J].传感技术学报,2006,19(6):2625-2629.
[15] Acharya D,Rani A,Agarwal S,et al.Application of adaptive Savitzky-Golay filter for EEG signal processing[J].Perspectives in Science,2016,8(C):677-679.
[16] 刘翔.BP算法的改进及其应用[D].太原:太原理工大学,2012.
[17] 张颖,李彩娟,邵惠鹤.基于主成分分析与模糊BP方法的藻类繁殖状态预测[J].上海交通大学学报.2012,46(5):785-789.
[18] 段洪涛,张柏,宋开山,等.长春市南湖叶绿素a浓度高光谱估测模型研究[J].水科学进展.2006,17(3):323-329.
[19] 何沁波.龙景湖叶绿素a浓度预测模型敏感性分析[D].重庆:重庆大学,2015.