Time Series Analysis using R
Instruction: Using R and the forecast package, demonstrate a monthly forecast with the built-in AirPassengers series. Train through December 1958 and validate on the next 24 months. Compare log-ARIMA with a seasonal-naive baseline using MAE, and explain time ordering, model selection and backtransformation limits.
Context: The candidate is required to explain the steps for conducting time series analysis in R. This should include data import, time series object creation, trend analysis, and forecasting. Emphasize on the use of specific R packages such as 'forecast' and functions like 'auto.arima' or 'ets'. This question tests the candidate's ability to handle time-series data, utilize R's packages for forecasting, and their understanding of statistical methods for time series analysis.
Updated
Official answer available
Read the opening below, then unlock the full answer and practical guidance.
First verify chronological ordering, sampling frequency, units and missing periods. ts(..., frequency=12) represents regular monthly observations; it does not inspect dates or repair an irregular calendar.
This R example uses the built-in positive monthly AirPassengers series and the forecast package...