The Trap of Reconstructed Data
When you have sparse or unreliable data, it often makes sense to intelligently fill in missing data with your best prediction, a process known as imputation. Of course, there can be multiple ways to fill in missing data. For example, let’s say we don’t know the average…