Given an indicator, for example "School enrollment, primary (% gross)" of the United States.
We have a series of discrete data from 1980 to 2009, which is counted for every years.
We want to predict the data for the year 2010 so we need to find a relation between it
and the previous data.
The problem is that I don't know which rule it is subjected to.
# Of course, the rules are different from indicators to indicators.
It is unlikely that we can predict the data for the year 2010 by simple (linear or polynomial)
interpolation or extrapolation.
I don't think that the data is subjected to some probabilistic rule like Markov chains.
My guess is that the data may be subjected to some Auto-Regression Moving-Average Models (ARMA) rule
or some time series rule like GARCH.
Being said that, I also have no idea which is the right direction to go or it is a no-go.
Please give me some hints,
Nguyen Vu Hung
- ► 2016 (22)
- ► 2015 (37)
- ▼ 2月 (6)
- ► 2008 (29)