Converts various data types to pandas datetime objects. Requires minimum "year", "month", "day" columns in DataFrame. Supports various input types including scalars, arrays, Series, and DataFrame
Moving average creates series of averages of different data selections. Mathematically viewed as low-pass finite impulse response filter. Used to smooth out short-term fluctuations and highlight long-term trends
Extrapolation estimates values beyond original observation range based on relationships. Similar to interpolation but with greater uncertainty and risk of meaningless results. Can extend methods or project known experience into new areas
Forecasting predicts future based on past and present data. Prediction is a more general term for forecasting. Risk and uncertainty are central to forecasting. Data must be up to date for accurate forecasts
Autocorrelation measures similarity between time series and its lagged versions. It uses same time series twice: original and lagged over successive intervals. Results range from -1 to +1, with +1 indicating perfect positive correlation
MSE measures variation between predicted and actual values. Calculated as average of squared differences between expected and actual values. Squaring ensures positive and negative differences don't cancel