Missing value Imputation in Bimodal Distribution

Hi all,

If the data-set has a bi-modal distribution. What would be the right imputation measure for missing values?

PS: This might be a beginner level question.


Hi @a.sethi,

Instead of using just one column, you can take help of another column and calculate the mode. For instance, if you have 2 columns job_type and gender, and you want to impute values in job_type which might have two modes. You can separately find mode of gender male and female of job_type, and then impute accordingly

© Copyright 2013-2019 Analytics Vidhya