0
$\begingroup$

I am doing a project that is looking at testing various imputation methods for estimating missing wind speed at a site in Trinidad. The dataset consist of hourly wind speeds for 15 years. I have tried simple methods like listwise deletion, mean substitution, linear regression, cublic & linear interpolation, nearest neighbor and even Markov chain monte carlo multiple imputation. Are there any other methods that could accurately estimate the wind speeds? I would also like to know if it is possible to use the respective Known wind speed data from a nearby site like in Tobago to replace the specific missing points in Trinidad?

$\endgroup$
  • $\begingroup$ This type of imputation might be reasonable if the wind speed in nearby areas are highly positively correlated with the wind speed at the missing location, $\endgroup$ – Michael Chernick Feb 23 '18 at 0:21
  • $\begingroup$ when i tested the correlation between the two sites, I got a value of 0.635. $\endgroup$ – L N Feb 23 '18 at 0:48

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.