I have data, that describes distance from given location to nearest object (e.g. school, shop etc). Because of performance reasons I couldn't scrape the data about objects, that are futher away than 2.5km. Now I have quite a lot of missing values in different variables).

My question is: What is the best way to fill such a missing data?

I considered such a solutions: - Binning data into few categories - one of them would be "more then 2.5km", - Replacing missing data with maximum value of each variable, - Replacing missing data with value greater then maximum.

Disadvantage of the first solution is, that I lose a lot of information. The other solutions disturb the distribution of the variables.

I use those variables in xgboost model both regression and classification.

Could you please give me some advice, which solution of these would be better. Or maybe I am missing something and there is any other better solution? Scraping it again with greater radius is not a solution, because it will leave missing values again.


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.