Question 1: Let's say you are an Analyst at a large national retail store. The retail store offers paid memberships at 99 dollars a year. The store has decided to start same-day delivery for its paying members in Atlanta. The plan is to start with ZIP codes with a high concentration of paying members, and then expand to the rest of the city over time. Youve been asked to provide a list of ZIP codes with a high concentration of paying members. What are your thoughts on this approach?

Question 2: The UCI Adult Dataset (Links to an external site.) is a public data which lets you predict whether income exceeds $50K/yr based on census data. One of the features available in the dataset is work hours per week. Youll find that males on average have a higher annual income than females in this dataset because on average males work more hours than females per week. Should you consider work hours per week in your prediction model, or will it be considered discrimination?

Question 3: You are an Analyst in a large Healthcare organization. Your organization has recently launched a high-risk care management program. The idea is to predict future health needs of patients, and then for those with the highest predicted health needs, offer a care management program. Your organization decides to use health care cost as a proxy measure for health needs. What are your thoughts on this Label proxy measure?