Journal of Spatial Analysis Environmental Hazards
تحلیل فضایی مخاطرات محیطی
Journal of Spatial Analysis Environmental Hazards
Literature & Humanities
http://jsaeh.khu.ac.ir
1
admin
2423-7892
2588-5146
10.61186/jsaeh
fa
jalali
1400
9
1
gregorian
2021
12
1
8
3
online
1
fulltext
fa
مقایسه کارآمدی چهار روش هوش مصنوعی در پیش بینی خشک سالی
Comparison of the effectiveness of four artificial intelligence methods in predicting drought
تخصصي
Special
پژوهشي
Research
خشکسالی یک اختلال موقتی است که خصوصیات آن از منطقه­ای با منطقه دیگر متفاوت است، از این رو نمی­توان تعریف جامع­ و مطلق برای خشکسالی بیان نمود.در تحقیق حاضر، به منظور معرفی یک روش مناسب جهت پیش­بینی خشکسالی برای یک ماه آتی، چهار روش هوش مصنوعی شامل یادگیری عمیق (<span dir="LTR">Deeplearning</span>) (از شبکه الکس­نت که یکی از شبکه­های کانولوشن می­باشد استفاده شده است)، الگوریتم <span dir="LTR">K</span> نزدیک­ترین همسایه (<span dir="LTR">KNN</span>)، ماشین برداد پشتیبان چند طبقه (<span dir="LTR">SVM-MultiClas</span><span dir="LTR">s</span>) و درخت تصمیم (<span dir="LTR">Decision Tree</span>) در نظر گرفته شد. داده­های بارندگی 11 ایستگاه سینوتیک استان یزد طی دوره ­آماری 29 ساله (1988 تا 2017) به صورت ماهانه به عنوان داده­های آزمایشی مورد استفاده قرار گرفتند. شاخص بارش استاندارد شده <span dir="LTR">(</span><span dir="LTR">SPI</span><span dir="LTR">)</span> برای نشان دادن وضعیت خشکسالی از نظر شدت و مدت در مقیاس­های زمانی 1، 3، 6، 9، 12 و 24 ماهه محاسبه گردید. در ابتدا داده­های بارش به عنوان ورودی شبکه­های عصبی و کلاس­بندی <span dir="LTR">SPI </span> به عنوان خروجی شبکه­ها قرار داده شد. 80 درصد داده­ها برای آموزش و20 درصد داده­­ها برای تست شبکه­ها به کار گرفته شد. نتایج نشان داد که تمامی شبکه­ها توانایی پیش­بینی خشکسالی را داشته­اند، بر اساس معیار ارزیابی <span dir="LTR">macro-f1</span> شبکه <span dir="LTR">Deeplearning</span> در مقیاس زمانی 1 ماهه با 71/22 درصد، ناکارآمدترین روش و <span dir="LTR">Decision Tree</span> با 65/64 درصد، کارآمدترین روش بوده­اند، اما با افزایش مقیاس زمانی، شبکه <span dir="LTR">Deeplearning</span> عملکرد خود را بهبود بخشید، به­طوریکه در مقیاس زمانی 24 ماهه با 35/65 درصد، بهترین عملکرد مربوط به شبکه <span dir="LTR">Deeplearning</span> و بعد از آن، شبکه <span dir="LTR">SVM</span><span dir="LTR">-</span><span dir="LTR">MultiClass</span> با 40/57 درصد، قرار گرفت.
<strong>Comparison of the effectiveness of four artificial intelligence methods in predicting drought</strong><br>
<strong>Abstract</strong><strong><span dir="RTL"></span></strong><br>
<strong>P</strong><strong>roblem statement:</strong><br>
Drought is a temporary disorder whose characteristics vary from region to region, therefore, it is not possible to define a complete and absolute definition of drought. Drought is one of the most important natural disasters that can occur in any climate regime. Since drought is unavoidable, it is important to know it in order to optimally manage water resources. Drought prediction can play an important role in managing this phenomenon. In other words, recognizing and predicting this phenomenon is one of the topics of interest for scientists who are interested in solving the problem of water shortage. More than 80% of Iran's area is covered by arid and semi-arid regions and lack of rain is a predominant phenomenon in this region. So far, several methods have been proposed to predict drought. Each method offers different results in specific conditions. Therefore, identifying the best method for predicting drought in the climatic conditions of central Iran is essential.<br>
<br>
<strong>Material and methods:</strong><br>
In this research, in order to introduce a suitable method for predicting drought for the next month, four methods of artificial intelligence including Deeplearning (using the Alexnet network, one of the convoluted networks), K nearest neighbor algorithm (KNN), multi-class Support vector machines (SVM-MultiClass) and decision tree have been used. Monthly rainfall data from 11 syntactic stations of Yazd province during the 29-year statistical period (1988 to 2017) were used as experimental data. Standardized precipitation index (SPI) was calculated to indicate drought status in terms of severity and duration on 1, 3, 6, 9, 12 and 24 month time scales. Precipitation data was used as neural network input and SPI classification as network output and 80 percent of the data was used for training and 20 percent for testing the networks.<br>
In this study, the Recurrence Plot method was used to interpret the time series to convert these series into images and RG and B pages were created. To convert rainfall data into photos at 1, 3, 6, 9, 12 and 24 month time scales, photo layers were created using a standardized rainfall formula, and by merging these three output layers into colored photos and black and white photos were obtained. Using three pages created in MATLAB software and merging them, the output was in the form of a photo, which was placed as the input of the Alexnet network. Combination of Recurrence Plot to create images and deep learning network for classification of drought data has been used for the first time in this research. To evaluate the effectiveness of the classification strategy, standard criteria of accuracy, micro-F1 and macro-F1 were used.<br>
<br>
<strong>Results Description and interpretation:</strong><br>
The results showed that all networks were able to predict drought. However, on short time scales such as 3 and 9 months, the accuracy assessment criteria for some classes are zero and the methods of learning from these classes are practically ignored due to their lack of data. But on a larger time scale, this issue has been addressed and the data of those classes are well categorized. Deep learning network with image input could not predict well in the short term due to lack of data, but in the long term due to increased data has improved its performance and had the best performance. The SVM method at different time scales has shown unreliable and variable behaviors that can not be said to be a suitable method for predicting drought at different time scales. Decision Tree and KNN methods have been able to predict drought better in the short term than in the long term. The two methods have been closely related. .Based on the Deeplearning network macro-f1 evaluation criterion, the one-month time scale with 22.71% was the most inefficient method and the Decision Tree with 64.65% was the most efficient method, But with the increase in time scale, the Deeplearning network improved its performance, so that at the 24-month time scale with 65.35%, the best performance for the Deeplearning network followed by the SVM-MultiClass network with 57.40%. For future research, it is suggested that Decision Tree and KNN methods be used to predict short-term drought. In this study, with increasing the time scale and increasing the data used, these two methods have lost their effectiveness compared to the short term.<span dir="RTL"></span><br>
<br>
<strong>key words</strong>: Drought, Standardized Precipitation Index, Artificial Intelligence, Deep Learning, Alexent, Recarence Plot<span dir="RTL"></span><br>
خشکسالی, شاخص بارش استاندار شده, هوش مصنوعی, یادگیری عمیق, الکسنت, ریکارنس پلات
Drought, Standardized Precipitation Index, Artificial Intelligence, Deep Learning, Alexent, Recarence Plot
139
156
http://jsaeh.khu.ac.ir/browse.php?a_code=A-10-1072-1&slc_lang=fa&sid=1
Laleh
Sharifipour
لاله
شریفیپور
lidasharifipoor@gmail.com
10031947532846009603
10031947532846009603
No
Ardakan University
دانشگاه اردکان
Mohammad-Javad
ghanei-Bafghi
محمد جواد
قانعی بافقی
mjghaneib@ardakan.ac.ir
10031947532846009604
10031947532846009604
Yes
Ardakan University
دانشگاه اردکان
Mohammad Reza
kousari
محمد رضا
کوثری
mohammad_kousari@yahoo.com
10031947532846009605
10031947532846009605
No
Soil Conservation and Watershed Management
پژوهشکده حفاظت خاک و آبخیزداری
Ssan
Sharifipour
ساسان
شریفیپور
sasansharifipour@gmail.com
10031947532846009606
10031947532846009606
No
Malek Ashtar University of Technology
دانشگاه صتعتی مالک اشتر