Federated learning improves COVID-19 forecasts across regions

Researchers have developed a federated learning framework to improve COVID-19 forecasting across different regions, even when outbreaks are asynchronous. By training Long Short-Term Memory networks using data from Guangdong, Macau, and Portugal, the federated model achieved forecasting accuracy comparable to centralized training while preserving data privacy. The study demonstrated that leveraging transmission dynamics from earlier outbreaks in one region could enhance predictions in regions experiencing later outbreaks. AI

IMPACT Federated learning offers a privacy-preserving method for collaborative AI model training, potentially applicable to other cross-regional data challenges.

RANK_REASON The cluster contains an academic paper detailing a novel application of federated learning for epidemic forecasting. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• Chinese Center for Disease Control and Prevention
• COVID-19
• federated learning
• Flower
• Guangdong
• long short-term memory
• Macau
• Omicron
• Portugal

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →