How to fill missing values for "Credit History" in Loan Prediction Problem

loan_prediction

#1

Regarding Loan Prediction 3
how can I fill values in credit history, It is symmetric with all other variables except loan status, but, we can’t fill it with the help of loan status, so How can I handle this problem?


#2

Hey
You can fill credit history using any central tendencies. You can also find a relationship of credit history with other variables using bivariate analysis and fill it accordingly.


#3

As You can observe that in the data set Credit_History has only two values 1 and 0 and it has been be off default numeric data type.

One thing you can do is to convert it into factor then replacing the missing values with mode of Credit_History.

Another option is to check its correaltion with others and then imputate the missing value by either mean or median.

Another option is to use the missForest function or like it any other package which can prefix the missing values.

Another option could be that you apply RF / Xgboost without imputaution and then observe the VarImp / Feature Plot respectively and if these makes Credit_History as Imp feature then do the imputation or else can delete the missing values…

So basically try and apply all the theory and method that is there and if u come up with something new then share it here .