CO₂ Emissions Forecasting
The analysis focuses on CO₂ atmospheric concentrations dataset. It covers data preparation and exploration using visualization and STL decomposition, checking stationarity by applying the Dickey-Fuller test and computing and plotting the ACF and PACF values. I also built a seasonal ARIMA model to produce one-step ahead, dynamic and future forecasts for CO₂ emissions forecasting.
Using grid search were identified the best parameters for building a seasonal ARIMA model. The model was used for producing one-step ahead, dynamic, and future forecasts.
The future forecasts indicate that the CO2 time series is expected to continue increasing.