A project on forecasting SARS-CoV-2 concentrations in wastewater at State College.

Data Used: NWSS Public SARS-CoV-2 Wastewater Metric Data

Provided By: National Wastewater Surveillance System

Tools Used:

• Time Series Analysis: Statsmodel, Facebook Prophet (Forecasting at Scale)
• Machine Learning Development and Packaging: mlflow
• Containerization: Docker
• Continuous Integration/ Continuous Development(CI/CD): Jenkins

Project Link

Results:

• Data Plot: Plot of SARS-CoV-2 concentrations in wastewater across two years

• Timeseries Decomposition: Season-Trend decomposition of the timeseries data

• Train Test Split: Splitting the data for training and testing

• Prophet Forecast: Prophet forecast with actual observations, changepoints and comparison with naive forecast

• Model Components: Time series components learnt by Prophet Model