Bias Is a Big Problem in Decision-Making, But So Is "Noise"

Bias has emerged as an increasing concern in machine learning (ML) and other analytic models. Let's take a look at the definition of bias. In a recent New York Times article, Daniel Kahneman, Olivier Sibony, and Cass R. Sunstein assert that bias is an error that causes people to have inclinations or opinions toward or against something. This is why we worry so much that the historical data used may create a predictable error in an analytic or machine learning model. But beyond bias is the issue of "noise" in decision-making.