Zeros in the target data
published 23 Mar 2020, 16:18

If you have much zeros at the target in a linear regression problem how do you handle the problem eq. 0 to 1 are target variables but 0 are about 70%

What will I do


I am just delving into dat analytics and therefore not such an authority. If I get this right, you have about 70% of zeros for your target variables than they are 1s. to me this is not a problem but an acceptable outcome since its justthe way the data is. Assuming this is the case and I was required to give a report on my data, I'd have it as " 70% of my population do not subscribe to this school while 30% do".


I mean in the case of continuous variables eg 0.8,0,0.03,0.2 and you are having 70% of 0 in the target column

How do you go about the zeros since they are too much to compare to other numbers

You get