Recognition of biases in data and models

A method of training a predictive model to predict a likely field value for one or more user selected fields within an application, including analyzing a pre-existing user-provided data set for biases.