In a world increasingly vulnerable to infectious disease outbreaks, the importance of accurate forecasting tools has become critical. Rising to meet this challenge is Dr Yisa Adeniyi Abolade, a mathematical statistician at Georgia State University, whose research is setting a new standard in epidemic modelling.

His recent study, co-authored with Dr Yichuan Zhao, introduces a refinement to the SEIR (Susceptible-Exposed-Infectious-Recovered) model, a cornerstone of epidemiological forecasting. The innovation lies in the application of the Sum of Absolute Deviations (SAD) method—a robust estimation technique that significantly improves the reliability of predictive models under real-world data conditions.

“The issue with traditional methods like least squares is that they are very sensitive to outliers,” said Abolade. “In public health data, irregularities are the rule rather than the exception, and our models need to be strong enough to handle that.”

By coupling SAD with a dynamic β(t) transmission rate—capable of reflecting seasonal effects or shifts in policy—the model becomes not only more resilient but also better aligned with the chaotic nature of disease spread. The research tested various functional forms for β(t), such as cosine and exponential patterns, mimicking how real-world conditions fluctuate.

“Our findings showed that SAD-equipped models outperformed traditional methods, especially in short-term forecasts,” Abolade explained. “This is crucial for public health officials who need dependable information quickly during emergencies.”

The study applied both simulated and real infectious disease data, demonstrating how the SAD method yielded more stable parameter estimates and reduced the risk of distortion by anomalous data points.

This work doesn’t only represent theoretical progress. It has direct implications for improving public health interventions, especially in regions where data may be incomplete or inconsistent. “We need models that can adapt to the messiness of real life,” said Abolade. “That’s the only way we can turn numbers into meaningful action.”

Dr Abolade’s background in statistical analysis and public health preparedness makes him a vital contributor in this area. With a focus on real-time forecasting, he envisions a future where epidemic response strategies are guided by smarter, more adaptive tools.

“There is a growing demand for models that don’t just simulate but respond, evolve and correct themselves,” he noted. “What we’ve created is a blueprint for that next generation of disease forecasting.”

The implications of this study are far-reaching. Health ministries, research bodies and global organisations could benefit from integrating SAD-enhanced SEIR models into their epidemic preparedness systems. The improvements are especially relevant in low-resource settings where perfect data is often unavailable.

Abolade believes that the greatest strength of the SAD approach is its simplicity. “It’s a modest change in method but makes a huge difference in outcomes. That’s the kind of innovation I aim for—practical, accessible and impactful.”

With public health emergencies expected to become more frequent due to climate change and urbanisation, research like this is not only timely but vital. It equips global health systems with sharper tools and helps reduce response time during critical moments.

For Abolade, this research is part of a broader mission to connect academic statistics with public service. “When our models make a difference in policy or healthcare delivery, that’s where our work finds its true value,” he said.