Back to problems

Predict bike demand and avoid overfitting

System Design · Two Sigma · Medium

Assume you have access to a bike-sharing service's operational records. The dataset aggregates, for each docking station and hour, fields such as the station identifier, counts of bicycles taken and returned, total dock capacity, number of bikes present at the beginning of the hour, meteorological data, holiday flags, and indicators of nearby public-transit activity or special events. Your assignment is to build a predictor that estimates how many bikes will be checked out…

Checking your access…