Predict fan engagement with baseball player digital content

You are tasked with forecasting four different measures of engagement (target1-target4) for a subset of MLB players who are active in the 2021 season. The data contains a set of static files that do not change with time (players.csv, teams.csv, seasons.csv, awards.csv) as well as daily data (train.csv) which is grouped by day. When predicting on a given date, you are forecasting the target variables for the next day (i.e. for date d, you're predicting the engagement for day d+1).