What is multivariate time series forecasting and how does it differ from univariate forecasting?

Multivariate forecasting uses multiple related time series as inputs, while univariate uses only one series.

Univariate forecasting models a single time series using its own past values. Multivariate forecasting uses multiple interdependent variables, capturing cross-series relationships. Models like VAR (Vector AutoRegression) and deep learning approaches like LSTMs handle multivariate dependencies effectively.