List All Projects List Graduate Projects List Senior Design Projects List REU Projects Other Undergraduate Projects
Project # 39
Classifying and Forecasting H5N1 Avian Influenza Across The Conterminous United States
| Team Lead: | Will Maberry |
| Other Team Members: | |
| Faculty Advisor(s): | Prasanna Gowda |
| Department: | CSE |
| Project Type: | REU |
| Title: | Classifying and Forecasting H5N1 Avian Influenza Across The Conterminous United States |
| Abstract: | Highly pathogenic avian influenza (HPAI) has caused widespread disruptions in the US poultry industry, backyard flocks, and avian wildlife. Despite recent seasonal HPAI recurrence and associated socioeconomic and ecosystem disruptions, there are no publicly available HPAI risk prediction tools or products available for the conterminous United States (CONUS). Here, we present two gradient booster machine learning ensemble models: 1, a classification model per month per county classifying positive (detected) or negative (non-detected) HPAI occurrence across the CONUS from June 2023 through January 2025; and 2, a forward-looking avian influenza risk (FLAIR) model providing low-, medium-, and high-risk classifications one month in advance per county per month across the CONUS from June 2023 through January 2025. The classification model's performance was balanced accuracy and recall of 0.78 and 0.83, respectively. The FLAIR model balanced accuracy and recall, overall, was 0.71 and 0.66, respectively. Classification model and FLAIR model feature importance corroborated with prior HPAI spatial epidemiological modeling and field studies, highlighting the roles of land use-land cover and climatological feature importances in HPAI outbreaks. The FLAIR monthly forecast has immediate, supportive applications towards a proactive HPAI risk management approach in US poultry and non-wild bird management in the CONUS. |
| Poster: | View Poster #39 (PDF) |
| Judges Scoring: | Submit Evaluation (judges only) |