弱弱的一个问题ds take home challenge里面的

处理missing value in your input variable时， predict the missing values using other variables， then include also those other variables and use a model that works well with correlated variables, like tree-based models for instance. 作者说虾面这句话是啥意思呢？Using the other variables your model would still be able to extract country-related informations + it could extractinformation from the fact that the user chose to not provide that information.
没看懂用这种方法处理的问题是啥？
跪求大神指点

感觉是用其他variable来做null处理，然后因为处理后的数据是predict出来所以和用来predict的input有相关性，所以需要用tree based model来做最后的预测，因为tree base对数据本身相关性不敏感。最后那句话意思是用这种方法处理null的优势是可以解决比如有些null是合理的，用户没有提供这个variable