[论文解读] Coronavirus (COVID-19): ARIMA based time-series analysis to forecast near future

Hiteshi Tandon, Prabhat Ranjan+2 more|arXiv (Cornell University)|Apr 16, 2020

COVID-19 epidemiological studies参考文献 5被引用 76

一句话总结

本文开发了基于 ARIMA 的时间序列模型，用于预测 India 的近未来 COVID-19 病例，指出一个上升且指数增长的趋势，以帮助规划。

ABSTRACT

COVID-19, a novel coronavirus, is currently a major worldwide threat. It has infected more than a million people globally leading to hundred-thousands of deaths. In such grave circumstances, it is very important to predict the future infected cases to support prevention of the disease and aid in the healthcare service preparation. Following that notion, we have developed a model and then employed it for forecasting future COVID-19 cases in India. The study indicates an ascending trend for the cases in the coming days. A time series analysis also presents an exponential increase in the number of cases. It is supposed that the present prediction models will assist the government and medical personnel to be prepared for the upcoming conditions and have more readiness in healthcare systems.

研究动机与目标

• 需要预测近未来的 COVID-19 感染，以支持防控和医疗规划。
• 开发一个基于 ARIMA 的 India COVID-19 病例预测模型。
• 预测未来的病例数量，以协助政府和医疗就绪。

提出的方法

• 构建并应用一个基于 ARIMA 的时间序列模型来 COVID-19 病例数据。
• 使用所开发的模型预测 India 的近未来感染数量。
• 评估并解释预测，以了解增长轨迹（上升/指数）。
• 提出对卫生系统准备的潜在影响。

实验结果

研究问题

• RQ1 在 ARIMA 模型捕捉下，India 的近未来 COVID-19 病例轨迹是什么？
• RQ2 基于 ARIMA 的预测是否能揭示 India 的 COVID-19 病例呈上升甚至潜在的指数增长模式？
• RQ3 这些预测如何为政府和医疗保健在未来条件下的准备工作提供信息？

主要发现

• 预测表明在不久的将来 COVID-19 病例呈上升趋势。
• 时间序列分析表明病例数量呈指数增加。
• 该模型旨在帮助政策制定者和医务人员进行准备和资源规划。