干旱区生态环境知识资源中心(Arid): Yield prediction of apricot using a hybrid particle swarm optimization-imperialist competitive algorithm- support vector regression (PSO-ICA-SVR) method

Arid

DOI	10.1016/j.scienta.2019.108756
	Yield prediction of apricot using a hybrid particle swarm optimization-imperialist competitive algorithm- support vector regression (PSO-ICA-SVR) method
	Esfandiarpour-Boroujeni, I 1; Karimi, E.2; Shirani, H.1; Esmaeilizadeh, M.3; Mosleh, Z.4
通讯作者	Esfandiarpour-Boroujeni, I
来源期刊	SCIENTIA HORTICULTURAE
ISSN	0304-4238
EISSN	1879-1018
出版年	2019
卷号	257
英文摘要	Determining the important factors that affect the crop yield would be helpful for optimization of irrigation water usage, fertilizer applications, and other inputs and resources in farms. The main objective of this study was to evaluate the performance of a hybrid particle swarm optimization-imperialist competitive algorithm-support vector regression (PSO-ICA-SVR) method to predict apricot yield and to identify important factors affecting its yield in Abarkuh region, Yazd Province, Iran. One hundred ten apricot orchards were selected randomly and soil samples were taken at two layers (0-40 and 40-80 cm). Besides, water samples and leaves from branches without fruit were taken in each sampling point. Management information and apricot yields were achieved by completing a questionnaire. After performing yield modeling in MATLAB software, the results showed that the application of this hybrid algorithm identified 18 variables out of 61 studied variables as the most effective variables in apricot yield. The order of importance of selected variables in relation to the sensitivity analysis was irrigation intervals, spacing of rows planting, spacing of trees on each row, magnesium and nitrogen content of leaf, cultivar type, sand percentage and salinity of first studied layer, leaf sodium, soluble magnesium of first studied layer, sand percentage of second studied layer, age of trees, available potassium of first studied layer, silt percentage of second studied layer, coarse fragments percentage of first studied layer, soluble calcium, soluble magnesium and salinity of second studied layer. The validation results showed that the hybrid algorithm was able to estimate apricot yield with a relatively high accuracy (RMSE = L737 for training data and RMSE = 2.329 for testing data). Also, the use of efficiency index (EFI) showed that the hybrid algorithm has a high efficiency (about 99%) for prediction of apricot yield. As the apricot yield has a high sensitivity to the irrigation frequency, adopting management practices should be consider to increase water use efficiency in arid and semi-arid regions.
英文关键词	Crop yield modeling Data-based models Evolutionary algorithms Feature selection
类型	Article
语种	英语
国家	Iran
收录类别	SCI-E
WOS记录号	WOS:000486103700067
WOS关键词	NEURAL-NETWORKS ; SALT-TOLERANCE ; CROP YIELD ; GROWTH ; SALINITY ; MACHINE ; IRRIGATION ; NITROGEN ; QUALITY ; IMPACT
WOS类目	Horticulture
WOS研究方向	Agriculture
EI主题词	2019-11-17
资源类型	期刊论文
条目标识符	http://119.78.100.177/qdio/handle/2XILL650/310911
作者单位	1.Vali E Asr Univ Rafsanjan, Fac Agr, Dept Soil Sci, Rafsanjan 7718897111, Iran; 2.Vali E Asr Univ Rafsanjan, Coll Agr, Dept Soil Sci, Rafsanjan, Iran; 3.Vali E Asr Univ Rafsanjan, Fac Agr, Dept Hort Sci, Rafsanjan, Iran; 4.AREEO, Soil & Water Res Inst, Karaj, Iran
推荐引用方式 GB/T 7714	Esfandiarpour-Boroujeni, I,Karimi, E.,Shirani, H.,et al. Yield prediction of apricot using a hybrid particle swarm optimization-imperialist competitive algorithm- support vector regression (PSO-ICA-SVR) method[J],2019,257.
APA	Esfandiarpour-Boroujeni, I,Karimi, E.,Shirani, H.,Esmaeilizadeh, M.,&Mosleh, Z..(2019).Yield prediction of apricot using a hybrid particle swarm optimization-imperialist competitive algorithm- support vector regression (PSO-ICA-SVR) method.SCIENTIA HORTICULTURAE,257.
MLA	Esfandiarpour-Boroujeni, I,et al."Yield prediction of apricot using a hybrid particle swarm optimization-imperialist competitive algorithm- support vector regression (PSO-ICA-SVR) method".SCIENTIA HORTICULTURAE 257(2019).