کاربرد شبکه عصبی مصنوعی و رگرسیون لجستیک در پیش‌بینی حضور علف‌های‌هرز در مزارع نخود دیم استان کردستان

نوع مقاله: مقاله پژوهشی

نویسندگان

1 دانشگاه فردوسی مشهد

2 مرکز تحقیقات کردستان

چکیده

به منظور مقایسه توانایی مدل‌های شبکه عصبی مصنوعی با رگرسیون لجستیک در پیش بینی حضور علف‌های‌هرز، آزمایشی در 33 مزرعه نخود دیم استان کردستان در سال زراعی 92-1391 انجام شد. برای این منظور، اطلاعات اقلیمی و خاکی به عنوان متغیرهای مستقل و حضور و عدم حضور علف‌های‌هرز غالب به عنوان متغیرهای وابسته در مدل‌های رگرسیون لجستیک و شبکه عصبی مصنوعی استفاده شدند. در این تحقیق از شبکه پرسپترون چندلایه با نه نرون در لایه ورودی، یک و دو لایه پنهان با تعداد نرون مختلف و دو نرون در لایه خروجی استفاه شد. علف‌هرز بی تی راخ (Galium aparine L.) و پیچک صحرایی (Convolvulus arvensis L.) با بیشترین شاخص فراوانی، علف‌های‌هرز غالب مزارع نخود بودند. نتایج نشاه داد که رگرسیون لجستیک نتوانست بین متغیرهای مستقل و حضور علف‌هرز بی تی راخ معادله ای را برازش دهد. در حالی که شبکه عصبی مصنوعی قادر بود برای هر دو علف‌هرز بی تی راخ و پیچک صحرایی در هر دو مرحله نمونه برداری، مدل مناسبی را برازش دهد. بطور کلی شبکه‌های عصبی مصنوعی با کارایی بالا در مقایسه با روش رگرسیون لجستیک، برای پیش‌بینی حضور علف‌های‌هرز در مزارع نخود دیم استان کردستان، مناسب تر بوده و کاربرد آن برای این منظور قابل توصیه می‌باشد.

کلیدواژه‌ها


عنوان مقاله [English]

Application of Artificial Neural Network (ANN) and Logistic Regression for Predicting Weeds Presence in Dryland Chickpea Fields of Kurdistan Province

نویسندگان [English]

  • Sahar Mansourian 1
  • Ebrahim Eizadi Darbandi 1
  • Mohammad Hassan Rashed Mohassel 1
  • Mehdi Rastgoo 1
  • Homayoun Kanouni 2
چکیده [English]

A survey was conducted to compare the potential of ANN and logistic regression in predicting weed presence of 33 dryland chickpea fields in Kurdistan province. Climatic and edaphic factors as independent variable and presence or absence of weeds with highest abundance as dependent variables were entered in the logistic regression and ANN models. The developed ANN was a Multi Layer Perceptron with nine neurons in the input layer, one and two hidden layer(s) of various numbers of neurons and two neurons in the output layers. Catchweed (Galium aparine L.) and field bindweed (Convolvulus arvensis L.) with the highest abundance indices were the dominant weeds in the chickpea fields. The logistic regression did not fit a model for catchweed, however, the ANN could develop the best suited models for predicting the catchweed and field bindweed presence in dryland chickpea fields. Sensitivity analysis revealed that altitude and rainfall are the most significant parameters in modeling weed presence in dryland chickpea fields. For the optimal model, the values of the model’s outputs correlated well with actual outputs and its application for this purpose is recommended.

کلیدواژه‌ها [English]

  • Abundance Index
  • modeling
  • Multi Layer Percepteron

(Chickpea, Lenze and Mung). Journal of Research and promotion. Agriculture Organization of Kurdistan province.16 pages. (In Persian with English Summary).

Aitkenhead, M.J., Dalgetty, I.A., Mullins, C.E., Mcdonald, A.J. S. and Strachan, N.J.C. .2003. Weed and crop discrimination using image analysis and artificial intelligence methods. Comp. & Elec. Agric. 39:157-171.

Alimoradi, L., Rashed Mohassel, M., Khazaee, H. and Ahmadian Iazdi, A. 2010. Weed community response to Sugar been field condition. 3rd Iranian Weed Science Congress: 148-150. (In Persian with English Summary).

Allison, P. D. 2001. Logistic Regression Using the SAS System: Theory and Application.Wiley Interscience, New York. 288 pp.

Anonymous, 2011. Almanac of agricultural and horticultural products. Statistic and information processing office of planning and economic of Ministry of Jahad- Agriculture, Tehran. (In Persian with English Summary).

Anonymous, .2007. Matlab. The Language of Technical Computing. Version 7.4 . The Mathworks of Natick, Massachusetts, USA.

Azadeh, A., S.F. Ghaderi and S. Sohrabkhani. 2006. A simulated-based neural network algorithm for forecasting electrical energy consumption in Iran. Energ. Policy. 36: 2637-2644.

Ayalew, L. and Yamagishi, H. (2005) The application of GISbased logistic regression for landslide susceptibility mapping in the Kakuda–Yahiko Mountains, Central Japan. Geomorphology. 65: 15–31.

Benvenuti, S. 2003. Soil texture involvement in germination and emergence of buried weed seeds. Agron. J. 95: 191-198.

Bridges, D. C., Brecke, B. J. and Barbour, J. C. 1992. Wild poinsettia (Euphorbia heterophylla) interference with peanut (Arachis hypogaea). Weed Sci. 40: 37-42.

Burk, A. 2001. Classification and ordination of plant communities of the Nauntain, Namibia. J. Veg. Sci. 12: 53-60.

Callaway, R. M., Thelen, G. C., Barth, S., Ramsey, P. W., and Gannonn, J. E. 2004. Soil fungi alter interactions between the invader Centaurea maculosa and North America natives.               Ecol. 85: 1062-1071.

Chau, K.T. and Chan, J.E. 2005. Regional bias of landslide data in generating susceptibility maps using logistic regression for Hong Kong Island. Original Article. 280-290 pp.

Dai, F.C., Lee, C.F. 2002. Landslide characteristics and slope instability modeling using GIS, Lantau Island, Hong Kong. Geomorphology. 42: 213–228.

Datta, A., Sindel, B.M., Jessop, R.S., Kristiansen, P. and Felton, W.L. 2007. Phytotoxic response and yield of chickpea (Cicer arietinum) genotypes with pre-emergence application of isoxaflutole. Aust Expe Agri. 47: 1460-1467.

FAO. 2011. FAO Year Book. FAO Publication.

Fried, G., Norton, L. R. and Reboud, X. 2008. Environmental and management factors determining weed species composition and diversity in France. Agriculture, Ecosyst & Enviro. 128: 68–76.

Goslee, C. S., Peters, D. P. C. and George-Beck, K. 2006. Spatial prediction of invasion success across heterogeneous landscapes using an individual-based model. Biological. invas. 8: 193- 200.

Haykin, S., 1994. Neural networks: A comprehensive foundation. McMillan College Publishing Company, New York.

Hassan, G., Khan,I., Khan,M.Z., Shah, N.H., Khan, M. and Liaqatullah, M. 2010. Weed flora of chickenpea in district Lakki Marwat, NWFP, Pakistan. Sarhad J. Agric. 26: 79-86.

Heidari, M.D. Omid, M. and Akram, A. 2011. Application of Artificial Neural Network for Modeling Benefit to Cost Ratio of Broiler Farms in Tropical Regions of Iran. Res. J. App. Sci. Eng. and Tech. 3(6): 546-552.

Hosmer, D. W. and Lemeshow, S. 2000. Applied logistic regression. Wiley, New York. 307 pp.

Hussain, F., Murad, A. and Durrani, M.J. 2004. Weed communities in wheat fields of Mastuj,         District Chitral, Pakistan. Pak. J. Weed Sci. Res. 10: 101-108.

ICARDA (International Center for Agricultural Research in the Dry Area-Farming System Program). Annual Reports. 2007. Aleppo, Syria.

Irmak, A., Jones, J.W., Batchelor W.D., Irmak, S., Boote, k.J. and Paz, j.O. 2006. Artificial neural network model as a data analysis tool in precision farming. Transactions of the   American Society of Agricultural and Biological Engineers. 49: 2027-2037.

Iswar, D., Sashikant, S., Cees, V.W., Alfred, S., Robert, H. 2010. Landslide susceptibility assessment using logistic regression and its comparison with a rock mass classification system, along a road section in the northern Himalayas, India. Geomorphology. 114: 627-637.

Kadmon, R. and Danin, A. 1999. Distribution of plant species in Israrl in relation to apatial variation in rainfall. J. Veg. Sci.10: 421-432.

Kaul, M., Hill, R.L. and Walthall, C. 2005. Artificial neural networks for corn and soybean yield prediction. Agriculture system. 85: 1-18.

Kavdir, S. 2004. Discrimination of sunflower, weed and soil by artificial neural networks. Comp. & Elec. Agric. 44: 153-160.

Khadem Alhoseini, Z., Shokri. M. and Habibian, S. H. 2007. Survey of topography and climate roles on plant distribution in rangeland of Mashjar Arsanjan (Bonab). J. Rang. 3: 222-235. (In Persian with English Summary).

Kropff, M.J. and Lotz, L.A.P. 1992.System approach to quantify crop- weed interaction and their application to weed management.Agri Sys. 40: 256 – 282

Liu, C., Berry, P. M., Dawson, T. P., and Pearson, R. G. 2005. Selecting thresholds of occurrence in the prediction of species distributions. Ecography. 28: 385-393.

Liu, J., Goering, C.E. and Tian, L. 2001. A neural network for setting target yields. Transactions of the American Society of Agricultural Engineers. 44: 705-713.

Magurran, A.E. 1988. Ecological Diversity and Its Measurement. Princeton University Press, Princeton, NJ, USA. 179 pp.

Mahmoudabadi, A. and Seyedhosseini, S. M. 2012. Time-risk tradeoff of hazmat routing problem in emergency situation. Proceedings of the tird international conference onindustrial engineering and operation management, Istanbul, Turkey: 344-351.

McCully, K.M., Simpson, G. and.Watson, A.K. 1991. Weed survey of Nova Scotia Lowbush (Vaccinilum angustifolium) fields. Weed Sci. 39:180-185.

Mennan, H. and Ngouajio, M. 2006. Seasonal cycles in germination and seedling emergence of summer and winter populations of catchweed bedstraw (Galium aparine) and wild mustard (Brassicakaber).Weed Sci. 54: 114–120.

Minbashi Moeini, M., Baghestani, M. A. and Rahimian Mashhadi, H. 2008. Introducing an abundance index for assessing weed flora in survey studies. Weed Bio. & Manag. 8:172-180.

Mortazavi Kooshk, N., Mirinejhad, Sh., Keshavarz, K. and Saeedi, K. 2011. Survey of altitude      effect on biological yield and morphological traits of spurge in Kohgiloieh and Boierahmad province. MSc. Thesis in Agronomy. Yasouj Branch. Islamic Azad University. (In Persian with English Summary).

Mousavi. S. K. and Ahmadi, A. 2009. Yield and its components of chickpea response to date and density of planting and interference of weed in dry land conditions of Lorestan province. Iranian J. Plant Protec. 23: 1-13. (In Persian with English Summary).

Murphy, C.E. and Lemerle, D. 2006. Continuous cropping systems and weed selection Euphytica 148: 61–73.

Niromand, H. 1995. Linear regression. Astan Ghods Razavi press. University of Imam Reza. (In Persian with English Summary).

Nordmeyer, H. and Dunker, M. 1999. Variable weed densities and soil properties in a weed              mapping concept for patchy weed control. Proceedings of the 2nd European Conference on Precision Agriculture, Odense Congress Centre, Denmark: 453-462.

Norouzi. Sh., D. Mazaheri. and Ghanbari, A. 2003. Servey of weed competition effects on yield and yield components of wheat in Shirvan area. Agro. J. (Pajouhesh & Sazandegi). 60: 91-96. (In Persian with English Summary).

Ozdemir, A. 2011. Using a binary logistic regression method and GIS for evaluating and mapping the groundwater spring potential in the Sultan Mountains (Aksehir, Turkey).      Journal of Hydrology. 405: 123–136.

Pearce, J., Ferrier, S. 2000. Evaluating the predictive performance of habitat models developed using logistic regression. Eco. Model. 133: 225-245.

Pepe, M.S. 2000. Receiver operating characteristic methodology. J. Am.                Statis. Ass. 95: 308-311.

Pinke, G., Pal, R. and Botta-Dukat, Z. 2010. Effects of environmental factors on weed species composition of cereal and stubble fields in western Hungary. Ent. Eur. J. Biol. 5: 283–292.

Pradhan, B. 2010. Manifestation of an advanced fuzzy logic model coupled with Geo-information techniques to landslide susceptibility mapping and their comparison with logistic regression modeling. Environ. Eco. & Statis. 23pp.

Rahmani. A. and Esmaili. Gh. 2010. Efficacy of Artificial Neural Networks, Logestic Regression in Necol prediction. Quartery of economic (Economic Survey). 4: 151-172. (In Persian with English Summary).

Rashed Mohassel, M.H., Najafi, H. and Akbarzadeh, M.D. 2009. Weed biology and control.           Ferdowsi University of Mashhad Press, Second Edition, Pp 404. (In Persian with English                 summary).

Reineking, B. and Schroder, B. 2006. Constrain to perform: regulation of habitat models.                Ecological Modelling. 193: 675-690.

Rohani. A. and Makarian, H. 2011. Preparing maps of weed management by Artificial Neural       Networks to apply in precise agriculture. J. Agri. Machi. Eng. 1: 74-83. (In Persian with English Summary).

Sanjari, S. 2011. Application of Arc GIS 10. Mehregan Ghalam Press.408 Pp.

Shimi, P. and Termeh, F. 1994. Weeds of Iran. Agriculture Research, Education and Extension Organization. 154 Pp.

Silc U., Vrbnicanin, S., Bozic, D., Carni, A. and Stevanovic, D. 2009. Weed vegetation in the north-western Balkans: diversity and species composition. Weed Res. 49: 602-612.

Thomas, A. G. 1985. Weed survey system used in Saskathevan for cereal and oilseed crops. Weed Sci. 33:34-43.

Thomas, A.G. and Donaghy, D. I. 1991. A survey of the occurrence of seedling weeds in spring annual crops in anitoba. Can. J. Plant Sci. 71: 811-820.

Torrecilla, J.S., Otero, L. and Sanz P.D. 2004. A neural network approach for thermal/pressure        food processing. Food Eng. 62: 89-95.

Tyrer. S.J., Hild. A.L., Mealor. B.A., and Munn. L.C. and Duncan. C.L. 2001. Knapweed management: another decade of change. Proceedings of the 1st international knapweed symposium of the twenty first century. 110 p.

Vakil-Baghmisheh M.T. 2002. Farsi Character Recognition Using Artificial Neural Networks. PhD Thesis, Faculty of Electrical Engineering, University of Ljubljana.

Walter. A.M., Christensen. S. and Simmelsgaard. S.E. 2002. Spatial correlation between weed species densities and soil properties. Weed Res. 42: 26-38.

Yang, C.C., Prasher, S.O., Landry, J.A. and Ramaswamy, H.S. 2003. Development of a erbicideapplication map using artificial neural networks and fuzzy logic. Agr. Sys. 76: 561-574.

Zanin, G., Otto, S., Riello, L., and Borin, M. 1997. Ecological interpretation of weed flora dynamics under different tillage systems. Agriculture. Ecos & Environ. 66: 177–188.

Zhang, W.J., X.Q. Zhong, and G.H. Liu. 2008. Recognizing spatial distribution patterns of grassland insects: neural network approaches. Stochastic Environmental. Res. & Risk Assess. 22:207–216.