time series analysis (TSA)

Definition

Trend forecasting (extrapolation) techniques (such as autoregression analysis, exponential smoothing, moving average) based on the assumption that 'the best estimate for tomorrow is the continuation of the yesterday's trend.' TSA is more suitable for short-term projections and is used where (1) five to six year's time series data is available and (2) where relationships between different values of a variable and their trend is clear and relatively stable. Instead of building a cause-and-effect (causal) model, TSA aims to isolate the sources of variations in a set of data so that their effect on a variable can be determined.