Time-series prediction/decomposition:

  1. Year-on-year growth;
  2. Annual pattern;
  3. Weekly pattern (weak);
  4. Daily pattern (strong);
  5. Events: holiday (esp. Thanksgiving and Christmas) and severe weather (hurricane, storm in summer/fall; blizzard in winter);
  6. Random noise unaccounted for;

No de-trending for non-count series: Intensive quantities e.g. probability density do not need de-trending (normalization).

Standard univariate time-series model, as implemented in R forecast package {Hyndman2008}. Multivariate time-series with black box model, as deployed in Uber {Laptev2017}.

X-13ARIMA-SEATS, a seasonal adjustment software developed by the United States Census Bureau.

Uncertainty in time-series models are often presented as 95% CI band, as in {Hanna2017}.