Bias in the model can happen through different stages of the modeling process. The most important stage is data selection and sometimes bias is the result of selection of the data itself, rather than an error with labeling the data.
A data scientist should spend some time and take different steps to ensure selected data has high quality and check for different types of bias that could potentially exist within the data. In this latest Data Science Central podcast we will talk about different categories of data bias and different approaches to avoid a bias decision as a result of low quality data.
Marzi Rasooli, Data Science Solutions Specialist for Financial Services - SAS Canada
Rafael Knuth, Contributing Editor - Data Science Central