fa ارزیابی و مقایسه کارایی کرنل های ثابت و انطباقی در رگرسیون وزندار جغرافیایی برای مدلسازی بیماری لپتوسپیروز در استان گیلان سامانه های اطلاعات مکانی GIS پژوهشي Research <p dir="RTL" style="text-align: justify;">بیش از 200 نوع بیماری مشترک بین دام و انسان در جهان وجود دارند که لپتوسپیروز یکی از مهمترین آنها به شمار می<span dir="LTR">&shy;</span>رود. هر ساله با شروع فصل شالیکاری در شمال کشور، این بیماری شیوع پیدا کرده و در مواردی منجر به مرگ و میر می<span dir="LTR">&shy;</span>شود. در مقایسه با سایر استانها در چند سال اخیر، استان گیلان بیشترین میزان گزارش بیماری لپتوسپیروز را به خود اختصاص داده است و از این جهت مطالعه و مدلسازی این بیماری در استان فوق از الویت زیادی برخوردار است. در این تحقیق کارایی کرنل<span dir="LTR">&shy;</span>های ثابت و انطباقی در روش رگرسیون وزندار جغرافیایی (<span dir="LTR">GWR</span>) جهت مدلسازی بیماری لپتوسپیروز و تاثیر 7 عامل محیطی بر آن در سال<span dir="LTR">&shy;</span>های 2009 تا 2011 در سطح دهستان<span dir="LTR">&shy;</span>های استان گیلان مورد بررسی قرار گرفته است. از دو معیار ضریب تبیین و خطای میانگین مربعات نیز برای ارزیابی نتایچ استفاده شده است. نتایج به دست آمده حاکی از آن است که کرنل انطباقی نسبت به کرنل ثابت و نیز وزندهی زوج مربعی در مقایسه با وزندهی گوسین به ترتیب عملکرد بهتری دارند. از نظر معیارهای انتخاب پهنای باند به ترتیب <span dir="LTR">AIC</span>، <span dir="LTR">CV</span> و &nbsp;<span dir="LTR">BIC</span>نقش مفیدتری دارند. از نظر متغیرهای محیطی نیز، پارامترهای متوسط دما، رطوبت و تبخیر بیشترین رابطه مثبت و ارتفاع و شیب رابطه عکس با بیماری لپتوسپیروز نشان می&shy;دهند. نقشه<span dir="LTR">&shy;</span>های تهیه شده از توزیع بیماری نشان می&shy;دهند که مناطق مرکزی استان گیلان بیشتر از سایر مناطق در معرض ابتلا به این بیماری بوده و مدیریت و کنترل این بیماری در این مناطق از الویت بیشتری برخوردار است.&nbsp;</p> <div style="text-align: justify;">There are more than 200 types of zoonotic diseases in the world and leptospirosis is the most important. Leptospirosis occurs mostly in areas with a tropical climate and abundant rainfall. There are no specific statistics of this disease worldwide, and records are underestimated for several reasons. Hence, the World Health Organization (WHO) named leptospirosis as a neglected tropical disease in the world and more research is needed in this field. When paddy season in the north of Iran begins, the disease spread and in severe cases leads to death. Leptospirosis is recognized globally as a multi-faceted disease and failure to recognize or treat it onetime can lead to death of patients. The main cause of the spread of this disease is a bacterium present in the body of domestic and wild animals, especially mice and dogs (as reservoirs of disease) and transmitted through the urine or feces to the environment. As a result, the bacteria can be transmitted to the human body through injuries to the skin or contact with contaminated soil and water. The environment and occupation are very effective in the spread of the disease, which is recognized as a work-related illness and can be dangerous in both urban and rural areas. The emergence of this disease can be due to reasons such as agriculture, livestock, butchers, recreational activities and water sports, poverty, travel to tropical areas, and any activity that leads to contact with water, soil or contaminated environment. This disease is more prevalent in fishermen and farmers, especially sugar cane farmers and workers, and it is very important to cause problems such as inability to work properly in the time and season needed planting and harvesting, as well as medical costs and even mortality. Compared to other provinces, Guilan province has the highest rate of leptospirosis recently. Therefore, the study and modeling of this disease in the province is of great importance. In this paper, the disease statistics in the rural area during 2009-2011 were assessed as the dependant variable and five variables considered as independent variables for modeling spatial distribution. Considering the important effects of bandwidth and weighting function on modeling results, the efficiency of fixed and adaptive kernels, Bi-Square and Gaussian weighting functions investigated. Two criteria were utilized to evaluate the results include MSE and Definition Coefficient. The results showed that the adaptive kernel and Bi-Square performed better than the fixed kernel and Gaussian, respectively. In terms of bandwidth selection criteria, AIC, CV and BIC played more meaningful role consecutively. Among the environmental variables, Temperature, Humidity and Evaporation illustrated positive relationship with disease and, elevation and slope showed negative relationship. The maps of the distribution of the disease indicated that the central regions of Guilan province are more prone to this disease than the other areas, and management and control of the disease in these areas is very important. Finally, all results were assessed by validation criteria and decision makers can use this helpful information for prevention programs and allocation of budget to the risky areas.</div> بیماری واگیر, لپتوسپیروز, سیستم اطلاعات مکانی, تحلیل مکانی, رگرسیون وزندار جغرافیایی Infectious Disease, Leptospirosis, GIS, Spatial Analysis, GWR 57 73 http://jgst.issgeac.ir/browse.php?a_code=A-10-421-2&slc_lang=fa&sid=1 A. Mohammadinia علی محمدی نیا Ali.mohamadinia@email.kntu.ac.ir 10031947532846005341 10031947532846005341 Yes K.N. Toosi University of technology دانشگاه خواجه نصیرالدین طوسی A. Alimohammadi عباس علیمحمدی alimoh_abb@kntu.ac.ir 10031947532846005342 10031947532846005342 No K.N. Toosi University of technology دانشگاه خواجه نصیرالدین طوسی Z. Ghaemi زینب قائمی zghaemi@mail.kntu.ac.ir 10031947532846005343 10031947532846005343 No K.N. Toosi University of technology دانشگاه خواجه نصیرالدین طوسی