-2

How to handle null values in dataset for performing feature scaling on a particular column?

That is to say, should we keep the null value as it is, or impute some value?

Is there any tutorial on how to handle null values while feature scaling?

blackraven
  • 5,284
  • 7
  • 19
  • 45
Pratik Goud
  • 1
  • 2
  • Data Science Stack Exchange: [Feature scaling data with missing values](https://datascience.stackexchange.com/questions/10343/feature-scaling-data-with-missing-values) – Adarsh Wase Jan 08 '22 at 07:53
  • 2
    I’m voting to close this question because it is not about programming as defined in the [help] but about ML theory and/or methodology - please see the intro and NOTE in https://stackoverflow.com/tags/machine-learning/info – desertnaut Jan 08 '22 at 12:07

1 Answers1

0

You can't do feature scaling when you have null values, you need to impute or drop the values.

Scaling:

  • It is a Scaling factor, it needs every element to scale individually. Ex: formula : data.mean - data ( assume ) # Scaling Formula

To scale all values in the data, we need every value to calculate mean as well as individual scaling factors also.

so Drop or Impute with other values.

Hope you understood!

Aravind R
  • 835
  • 1
  • 7
  • 14