Time-series forecasting models patterns such as trend, seasonality, cycles and changing relationships in sequential data. Applications include demand, energy use, finance, traffic and weather-related variables.
Forecast quality depends on the prediction horizon, available history, data shifts and uncertainty. Useful systems report ranges or probabilities rather than presenting every future value as certain.
ELI5
Time-series forecasting studies what happened over time and estimates what may happen next. The order and spacing of past observations are important.
For example, a retailer can use weekly sales history and seasonal patterns to estimate demand for the next month, while recognizing that an unexpected event may change the result.
