A data analytics approach for COVID-19 spread and end prediction (with a case study in Iran)

Arman Behnam, Roohollah Jahanmahin

This study, published in Modeling Earth Systems and Environment, introduces a comprehensive data analytics framework designed to model and predict the trajectory of the COVID-19 pandemic. Using Iran as a primary case study, the researchers move beyond simple case counts to introduce dynamic rates, providing a toolkit for evaluating intervention effectiveness and forecasting the eventual conclusion of an outbreak.

Methodology: A Multi-Tiered Analytics Approach

The researchers developed a three-tiered framework to process and interpret pandemic data, moving from historical description to future prediction. This systematic approach is essential for handling the noisy data often associated with novel pathogens.

The multi-stage data analytics framework for modeling and predicting COVID-19 trends.
  1. Descriptive Analytics: Historical data was used to calculate mortality and recovery rates relative to global means, establishing a baseline for Iran's performance.
  2. Prescriptive Analytics: The authors introduced the "Life Rate," a novel indicator calculated as the difference between the weekly recovery rate and the weekly death rate. A positive, increasing Life Rate serves as a pulse for the healthcare system's improving response.
  3. Predictive Analytics: The study applied Machine Learning (ML) and statistical functions to forecast future trends. Logistic functions were used to identify the "inflection point"—the moment growth begins to slow—while Gaussian functions were utilized for long-term trend fitting, showing superior performance in estimating the overall curve of the pandemic.

Key Insights: Life Rates and Demographic Vulnerability

A central finding of the research is that raw case numbers are less informative than the "Life Rate." By monitoring the net health of the response, policymakers can better understand the true state of the outbreak. Additionally, the study utilized the "Growth Factor" to identify when the pandemic moved from exponential growth to a deceleration phase (stabilizing around 1.0).

Demographic analysis confirmed a stark age divide in vulnerability. While the 35–65 age group represented the majority of non-recovered cases, mortality was heavily concentrated in the 55–85 age range. This reinforces the necessity for age-stratified protection policies rather than broad, uniform measures.

The analysis utilized countrywide daily information for Iran from February 19th to June 25th, 2020, aggregated from the WHO and Worldometer. The resulting models provided specific forecasts for the outbreak's progression.

Trends of confirmed, active, recovered, and death cases in Iran from March to August 2020.
  • Peak and End Forecasts: The model predicted the peak of the outbreak in Iran would occur approximately 150 days after onset, with a significant decline or "end" estimated between 290 and 320 days (around November 2020).
  • Quantitative Caps: Based on the trajectory at the time, the authors forecasted a maximum of roughly 200,000 confirmed cases and 10,000 cumulative deaths.
  • Model Robustness: Gaussian functions proved to be the most robust tool for long-term forecasting, as they naturally account for the rise and fall of infection waves more effectively than linear models.

Conclusion and Future Directions

While the framework successfully identified inflection points and forecasted trends, the authors acknowledge limitations regarding data granularity and the exclusion of external variables like humidity or solar radiation. Future research aims to integrate sociological and climatic parameters into ML algorithms to create more "environmentally aware" forecasting models that can better adapt to the complex realities of global health crises.

Summary updated

Abstract

World is now experiencing the new pandemic caused by COVID-19 virus and all countries are affected by this disease specially Iran. From the beginning of the outbreak until April 30, 2020, over 90,000 confirmed cases of COVID-19 have been reported in Iran. Due to socio-economic problems of this disease, it is required to predict the trend of the outbreak and propose a beneficial method to find out the correct trend. In this paper, we compiled a dataset including the number of confirmed cases, the daily number of death cases and the number of recovered cases. Furthermore, by combining case number variables like behavior and policies that are changing over time and machine-learning (ML) algorithms such as logistic function using inflection point, we created new rates such as weekly death rate, life rate and new approaches to mortality rate and recovery rate. Gaussian functions show superior performance which is helpful for government to improve its awareness about important factors that have significant impacts on future trends of this virus.
BibTeX
@inproceedings{behnam2021data,
  author = {Arman Behnam and Roohollah Jahanmahin},
  title = {{A data analytics approach for COVID-19 spread and end prediction (with a case study in Iran)}},
  booktitle = {Modeling},
  year = {2021},
  doi = {10.1007/s40808-021-01086-8},
}
Figure