What are the missing value imputation techniques?
The relaxed and quickest method to a missing data problem is dropping the offending entries. But we have to take care that dropping data not at random is dangerous and dropping too much data is also dangerous.
Dropping features with great nullity. A feature that has a high no of empty values is suspected to be very useful for prediction. It could frequently be safely dropped
The modest imputation technique is replacing missing values with the median or mean values of the dataset at large, or some similar summary statistic.