time series analysis[TSA]

time series analysis[TSA]　时间数列分析

Trend forecasting（extrapolation）techniques（such as autoregression analysis，exponential smoothing and 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 five to six year’s time series data is available where relationships between different values of a variable and their trend are 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.