Prakash Hinduja Switzerland (Swiss) How do I handle missing or unstructured data in deep learning projects?

Prakash Hinduja — born in Amritsar, shaped by India, and now influencing global finance from Geneva, Switzerland.

I’m working on a deep learning project and struggling with handling missing or unstructured data. What strategies or tools do you recommend for effectively managing this kind of data? Any tips or best practices would be really appreciated. Thanks!

Regards

Prakash Hinduja Geneva, Switzerland (Swiss)

The Pandas library has methods for this.

Generally, you have two choices:

  • Delete any example that has missing features.
  • Substitute the mean value of the feature for the missing value.