Scholarlink

<p>This study aims to analyse the trend of potato area, production, and yield in Russia and India by employing four different univariate time series models, i.e. ARIMA, Holt’s linear trend, alpha-ARIMA, and Sutte-ARIMA. Secondary data collected from the Food and Agriculture Organization (FAO) covering the period from 1992 to 2023 was utilized to forecast up to 2030. The comparative analysis represented that the α-ARIMA model was found to be the best in all three components of Russia and for area and yield in India, while the Holt linear trend model performed best for potato production in India, exhibiting maximum <i>R</i><sup>2</sup> values and minimum values of MAPE, MPE, RMSE, and MAE with a 95% accuracy level. The predicted results suggest that by 2030, Russia may experience continued growth in its area which would reach nearly 1151.27 thousand hectares with production estimated at 21,875.30 thousand tonnes, whereas yield is projected to decline up to 16,047.30&#xa0;kg/ha, while in India, potato production may increase nearly 67,036.34 thousand tonnes with moderate reductions in area and yield. Additionally, the instability analysis represented that Russia has experienced greater variability in recent years especially in area (11.2898) and production (15.072). The decomposition analysis further suggests that production growth in Russia was driven by yield whereas in India, the source of growth driven was shifted from yield to area. These findings provide valuable insights for policymakers, planners, and researchers, emphasizing the necessity for region-specific strategies to ensure the sustainability and resilience of potato production systems in both countries.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative Evaluation of Time Series Models for Decomposition and Forecasting of Potato Production Indicators in Russia and India

  • 期刊论文

  • 发表日期：

    2026年3月25日

查看全文
  • Aditya Bhooshan Srivastava,
  • Ashutosh Nayak,
  • Supriya,
  • Rishita Pandey,
  • Shikha Yadav,
  • Sanjith Bharatharajan Nair,
  • Vikash Pawariya,
  • Adelajda Matuka,
  • Yusra Tashkandy,
  • Walid Emam

摘要

This study aims to analyse the trend of potato area, production, and yield in Russia and India by employing four different univariate time series models, i.e. ARIMA, Holt’s linear trend, alpha-ARIMA, and Sutte-ARIMA. Secondary data collected from the Food and Agriculture Organization (FAO) covering the period from 1992 to 2023 was utilized to forecast up to 2030. The comparative analysis represented that the α-ARIMA model was found to be the best in all three components of Russia and for area and yield in India, while the Holt linear trend model performed best for potato production in India, exhibiting maximum R2 values and minimum values of MAPE, MPE, RMSE, and MAE with a 95% accuracy level. The predicted results suggest that by 2030, Russia may experience continued growth in its area which would reach nearly 1151.27 thousand hectares with production estimated at 21,875.30 thousand tonnes, whereas yield is projected to decline up to 16,047.30 kg/ha, while in India, potato production may increase nearly 67,036.34 thousand tonnes with moderate reductions in area and yield. Additionally, the instability analysis represented that Russia has experienced greater variability in recent years especially in area (11.2898) and production (15.072). The decomposition analysis further suggests that production growth in Russia was driven by yield whereas in India, the source of growth driven was shifted from yield to area. These findings provide valuable insights for policymakers, planners, and researchers, emphasizing the necessity for region-specific strategies to ensure the sustainability and resilience of potato production systems in both countries.