🟣 Terminology: Time Series Concepts

Stationarity: Statistical properties (mean, variance) don't change over time. Most models require this. Test: ADF (Augmented Dickey-Fuller). Fix: differencing.

Seasonality: Regular periodic patterns (daily, weekly, yearly).

Trend: Long-term directional movement (up, down, flat).

Autocorrelation: Correlation of a series with lagged versions of itself. High at lag 7? Weekly pattern.

ARIMA: AutoRegressive (past values predict future) + Integrated (differencing) + Moving Average (past errors predict future). Parameters: (p, d, q) = AR order, differencing order, MA order.