با همکاری انجمن اقتصاد کشاورزی ایران

خوشه‌بندی محصولات کشاورزی و دامی استان کردستان جهت توسعه زنجیره ارزش با استفاده از الگوریتم K-means

نوع مقاله : مقالات پژوهشی به زبان انگلیسی

نویسنده

گروه: گروه اقتصاد کشاورزی دانشکده: دانشکده کشاورزی دانشگاه کردستان

10.22067/jead.2026.99781.1438
چکیده
شناسایی نظام‌مند محصولات پرمزیت و هدایت هدفمند سرمایه‌گذاری در حلقه‌های تولید، فرآوری و بازاریابی، پیش‌شرط توسعه پایدار زنجیرههای ارزش کشاورزی است. استان کردستان علی‌رغم پتانسیل‌های قابل‌توجه تولیدی، با چالش‌های ساختاری نظیر شکاف میان تولید و فرآوری، بهره‌وری پایین و بالا بودن هزینه تولید محصولات کشاورزی مواجه است. ازآنجا که این موانع در سبد محصولات استان یکسان نیستند، سیاست‌گذاری واحد، ناکارآمد و موجب اتلاف منابع می‌شود. این پژوهش با هدف خوشه‌بندی محصولات کشاورزی و دامی استان کردستان بر اساس شاخص‌های کمی اقتصادی و ارائه چارچوبی کاربردی برای اولویت‌بندی سرمایه‌گذاری در توسعه زنجیره ارزش انجام شد. روش پژوهش توصیفی-تحلیلی با رویکرد داده‌کاوی است و بر پایه الگوریتم خوشه‌بندی کا-میانگین استوار است. برای این منظور، داده‌های مربوط به میزان تولید، عملکرد، هزینه‌های تولید، تقاضای سرانه، ارزش صادرات، ظرفیت واحدهای فرآوری و تعداد واحدهای فرآوری در استان برای ۱۶ محصول منتخب در بازه چهارساله ۱۴۰۰ تا ۱۴۰۳ جمع‌آوری و پس از استانداردسازی تحلیل شدند. شاخص‌های اعتبارسنجی سیلوئت، دان و دیویس-بولدین از انتخاب ساختار سه خوشه‌ای پشتیبانی کردند، سپس تحلیل واریانس یک‌طرفه و آزمون تعقیبی توکی برای تفسیر و پروفایل‌سازی خوشه‌های شناسایی‌شده به‌کار گرفته شد. خوشه اول شامل گندم و شیر با بالاترین حجم تولید و تقاضای داخلی اما کارایی صادراتی پایین؛ خوشه دوم شامل سیب‌زمینی، توت‌فرنگی، گردو، انگور، زردآلو و تخم‌مرغ با بالاترین ارزش صادرات، بهترین عملکرد و کم‌ترین هزینه تولید اما با شکاف جدی در زیرساخت‌های فرآوری؛ و خوشه سوم شامل هشت محصول با مقادیر پایین در اغلب شاخص‌های کلیدی. نتیجه کلیدی پژوهش، شناسایی تله خام‌فروشی محصولات پیچیده در خوشه دوم است. این طبقه‌بندی سه‌گانه، چارچوبی عملیاتی برای گذار از سیاست‌های یکنواخت به مداخلات تفکیک‌شده و اولویت‌بندی سرمایه‌گذاری در توسعه زنجیره ارزش استان فراهم می‌آورد.

کلیدواژه‌ها

موضوعات

عنوان مقاله English

Clustering of Agricultural and Livestock Products in Kurdistan Province for Value Chain Development Using K-means Algorithm

نویسنده English

p Alizadeh
Agriculture Faculty University of Kurdistan
چکیده English

Introduction

Systematic identification of advantageous agricultural products and the targeted direction of investment in production, processing, and marketing stages are prerequisites for the sustainable development of agricultural value chains. Kurdistan province possesses significant agricultural potential, contributing nearly 22.2 percent to the gross provincial product. It also holds a distinguished position in producing several key products, including wheat (ranked first in national rainfed production), potatoes (sixth), chickpeas (third), strawberries (first), and walnuts (second). Despite significant production potential, Kurdistan province faces structural challenges such as the gap between production and processing, low productivity and high costs of agricultural production. Consequently, a large portion of the province's agricultural output leaves the region in raw form, leading to inefficient utilization of comparative advantages and loss of added value. Since these barriers vary across the province's product portfolio, uniform policymaking is inefficient and wastes resources. This study aimed to cluster agricultural and livestock products in Kurdistan province based on quantitative economic indicators and provide a practical framework for prioritizing investment in value chain development.

Materials and Methods

The research method is descriptive-analytical with a data mining approach and is based on the K-means clustering algorithm. For this purpose, data on production, yield, production costs, per capita demand, export value, capacity of processing units, and number of processing units in the province were collected for 16 products during the four-year period 2021 to 2024 and analyzed after standardization. The K-means algorithm was selected for its high computational efficiency and accuracy in partitioning numerical data. The Silhouette, The Silhouette, Dunn, and Davis-Bouldin validation indices supported the selection of a three-cluster structure within the analyzed dataset. Subsequently, one-way ANOVA and Tukey's post-hoc tests were used to interpret and profile the identified clusters.

The findings from analyzing economic variables showed that agricultural and livestock products in Kurdistan province are highly dispersed across a wide range of indicators, with no single uniform pattern governing them. This multidimensional heterogeneity not only makes uniform policymaking and allocation of resources practically inefficient, but also prevents the accurate identification of key links in the value chain from production and processing to market and export in each of the products. Therefore, in order to transition from scattered description to systematic analysis and strategic targeting for value chain development, it is necessary to classify products into homogeneous groups based on structural similarities by using multivariate clustering methods. The results of multi indicator validation and statistical post hoc tests significantly confirmed the existence of three distinct patterns in the province's product portfolio. Cluster 1 included basic and high domestic demand products with high production volumes but low production efficiency and international competitiveness, which play a pivotal role in ensuring food security. Cluster 2, as the most important finding of the research, was identified as products with real competitive advantage in international markets, high yield, and low production costs; however, this group faces a deep gap in terms of processing infrastructure. Cluster 3 also included half of the province's products, which are at a lower level in terms of key economic and market indicators and have a limited production structure. The key finding of this research is the identification of a deep structural gap between production and processing in the province’s export-oriented products. Despite having significant production and market advantages, Kurdistan province loses a significant part of the added value of its products due to weaknesses in the complementary links of the value chain. This phenomenon, known in the theoretical literature as the trap of selling complex products in raw form, shows that production advantage alone does not guarantee economic development; rather, integrating it with processing, logistics, and market infrastructures is the primary prerequisite for enhancing value added. Overall, the present research showed that achieving sustainable development in the agricultural sector of Kurdistan province requires a transition from a quantitative production-oriented perspective to a qualitative value added perspective; and multidimensional clustering of products is an operational step in this direction.

Conclusion

Given the identified three-cluster pattern, policy making institutions in Kurdistan province should move beyond a uniform policy approach and develop a separate policy package for each cluster. Cluster 1 (wheat and milk), which are central to food security, the priority should be increasing productivity through modernization of machinery, development of contract farming, improved seeds, and stabilization of the domestic market through market regulation mechanisms. For cluster 2 (export oriented products with processing potential), which is the main driver of the province's value chain development, priority should be given to investing in processing infrastructure. This includes establishing specialized agro-processing zones for horticultural products, developing industries that convert grapes into raisins, juice, and concentrates, and processing walnuts and strawberries. Other priorities include completing the cold chain (by constructing controlled atmosphere cold storage facilities and deploying refrigerated transport fleets) developing sorting and packaging centers that meet international standards, facilitating access to target markets particularly Iraq and the Persian Gulf countries and supporting contract farming and regional branding. For Cluster 3 (products with limited economic advantage but a supportive role in food security), it is necessary to revise the cultivation pattern in favor of alternative products with higher comparative advantage, targeting subsidies to efficient producers, developing contract farming and production processing export consortia for products such as saffron, and maintaining special support for strategic food security products like chicken meat and fish.

کلیدواژه‌ها English

Advantageous products
Agricultural value chain
Clustering
K-means algorithm
ارسال نظر در مورد این مقاله
نام را وارد کنید.
نشانی پست الکترونیکی را به درستی وارد کنید.
وابستگی سازمانی را به درستی وارد کنید.
توضیحات را وارد کنید (حداقل 50 حرف)
CAPTCHA Image
شناسه امنیتی را به درستی وارد کنید.

مقالات آماده انتشار، پذیرفته شده
انتشار آنلاین از 02 آبان 1405

  • تاریخ دریافت 19 تیر 1405
  • تاریخ بازنگری 26 شهریور 1405
  • تاریخ پذیرش 02 آبان 1405
  • تاریخ اولین انتشار 02 آبان 1405