time series forecas

February 25, 2015


Report Abuse
Compute time series forecast for number of people entering a building daily
This experiment trains an R based auto-ARIMA model from forecast package, to predict how many people enter a building daily. The model uses Callt2 dataset from UCI Machine Learning repository. The data has been aggregated to daily counts of inflow, and the model assumes weekly seasonality. The scoring script returns point forecast for specified number of days into future.