COVID-19 Outbreak Prediction and Analysis using Self Reported Symptoms

It is crucial for policymakers to understand the community prevalence of COVID-19 so combative resources can be effectively allocated and prioritized during pandemic. Traditionally, has been assessed through diagnostic antibody testing data. However, despite increasing availability testing, required level not met in parts globe, introducing a need an alternative method communities determine disease prevalence. This further complicated by observation that spread vary across different spatial, temporal, demographic verticals. In this study, we study trends utilizing results self-reported symptoms surveys as complement reports. allows us assess prevalence, even areas with low ability. Using individually reported symptom data from various populations, our predicts likely percentage population tested positive COVID-19. We achieved mean absolute error (MAE) 1.14 relative (MRE) 60.40% 95% confidence interval [60.12, 60.67]. implies model +/- 1140 cases than original 1 million. addition, forecast location-wise next 30 days using previous days. The MAE 0.15 (MRE 11.28% [10.9, 11.6]) New York. present analysis these results, exposing clinical attributes interest demographics. Lastly, qualitatively analyze how policy enactments (testing, curfew) affect community.