How Do We Account For Biases When Creating Models And Selecting Data?

In the video below on CxO Talk, I asked Murli Buluswar, former Chief Science Officer of AIG and Michael Li, Chief Executive Officer of The Data Incubator about how to deal with biases in data.

In my view, data biases are a real threat. As models evolve, we have to make sure they don’t have biases:

  1. Prior to inclusion into a model
  2. During inclusion into a model
  3. After the recommendations by a model

Author: Khan

Speaker | Advisor | Blogger