I want to update the standard deviation of the future data to already calculated standard deviation of existing data using statistical modelling. This should work as per the behavior of deviation from the mean as we know the deviation tends to be a lot less as we see more and more samples tending towards mean. Any help or discussion is appreciated.
Take a look here: http://www.johndcook.com/blog/standard_deviation/.
There are many papers available on your topic. You find some rapid algorithms in Wikipedia: http://en.wikipedia.org/wiki/Standard_deviation
The true variance (i.e., variance in the population) can be predicted from the variance of a sample by simply multiplying the sample variance by N/(N-1), with N being the number of data points in your sample.
As you get more samples, you get various estimates of the true variance and from it you can predict the variance of a sample that you haven't got yet:
variance_of_future_sample=estimate_of_true_variance*(N-1)/N,
where N is that sample's size.
HTH.
Cheers,
SCM
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Statistical Modeling