Time series

Last updated: 21.03.2016

Time series represent a sequence of data for several time periods (hour, days, weeks, months, year). Independent variable is always time (axis x).  Time series is used to predict future and it has several components:

  • trend – represent long-term movement;  can be developed by regression analysis or moving averages; can be downward or upward
  • season – seasons repeat at defined short intervals (hours, days, weeks, months etc.) usually within a year, e.g. consumption of water in households is higher in the evening, consumption of gas is significantly higher during heating season (winter)
  • cycle – cycles usually last a few years and repeat in medium-term period, e.g. economic cycles
  • irregular (random) element – cannot be predicted, e.g. war


Pomohl Vám tento článek? Ohodnoťte jej prosím.

1 = nejhorší, 10 = nejlepší

1 2 3 4 5 6 7 8 9 10
Komentář

Related articles
Tento web používá k poskytování služeb, personalizaci reklam a analýze návštěvnosti soubory cookie. Používáním tohoto webu s tím souhlasíte. Další informace