Last Updated 01/28/2021

Mock MCM Problems

Instructions

Problem Statements

Problem B (Author: Amanda Johnson)

A new grocery store, Wai-Mart, is opening in your city. The owner is concerned with the slow checkout process at traditional grocery stores. He has hired your consulting team to propose two styles of stores and summarize the costs and benefits of each.

  1. Design a “traditional” grocery store with cashiers and checkout lines that will move customers through as efficiently as possible.
  2. Design a “smart” grocery store where customers are charged for items as they shop.
  3. Write a one-page memo to the owner of Wai-Mart summarizing your designs and the costs and benefits of each option.

 

Problem C (Author: Elise Bishoff)

Unbiased Financial Models

Background: Machine learning models have been found to be at risk of bias. This could mean unfair treatment to selected parties based on race, gender, etc. Financial institutions, like banks, have started adopting the use of machine learning models to help with automation, loan rates, loan applications, and detecting credit card fraud. Financial models are particularly in need of being unbiased, ethical, fair models. Otherwise, unfair models could potentially harm millions of people by denying them financial services that they would otherwise be qualified for had traditional processes been used.

Problem: You have been hired by Seattle Pacific Banks as a consultant. They are looking for a model that allows them to better decide whether to provide a loan to a given individual. They have asked you and your team to build a fair model that better predicts whether customers will pay back a loan or not.
The data file can be found here. These files contain loan data from the year 2017-2018. It includes a data dictionary about all variable types.  
 
Part I

A. Using the data provided, create a model that predicts the likelihood that someone will charge off on their loan or not.

B. Determine the best variables for predicting whether someone will pay back their loan in full or not.

Part II:

A. Financial models are subject to regulations to make sure they are fair and unbiased. This means that these models can’t discriminate against any parties based on race, gender, religion, etc. You have been tasked with making sure the model is fair. Define fairness for this use case. Create metrics for determine whether a model is fair.

B. Test whether you model from above is fair based off your metrics. Is there any bias hidden in your model?

C. Update model to better perform on your fairness metrics.

Part III:
Prepare a one-page memo to the Seattle Pacific Banks summarizing your definition of fairness, your fairness metrics, and how your model meets these requirements.

Your submission should consist of:

 

Problem C (Author: Wai and Amanda Johnson)

Your company has just received a contract from AARP to prepare rankings of hospitals in the Washington state. The rankings will be used to help AARP members to select hospital cares appropriated for their needs.

Your team is in charge of this project. Your supervisor would like you to
1. Look up recent performance data for hospitals in the Washington state.
2. Set up criteria/measures to establish the rankings appropriated to the needs of the AARP members. Carefully justify your measurement strategies.
3. Along with the rankings, put together a one-page memo to be sent to the AARP executives. In the memo, explains briefly how the rankings are being calculated in language that they can understand.

 

Problem E (Author Wai)

Bee Aware!
Asian giant hornet is the world's largest species of hornet. It attacks and destroys honey bee hives. In December 2019, Washington State Department of Agriculture (WSDA) received and verified two reports of Asian giant hornet near Blaine. These are the first-ever sighting in the United States. If it becomes established, this hornet will have negative impacts on the environment, economy, and public health of Washington State.

Your team has been hired as consultants for WSDA to provide projections for the potential spread of Asian giant hornet as well as the effects of different controlling measures on its population and spread.

1. Based on what we know about the insect, and the location of its last sighting, create mathematical models to give short-term and long-term projections of the spread of the insect.

2. Incorporate potential controlling measures into you model(s) and discuss the effects on

(a) the spread of the insect;

(b) the amount of projected negative impacts;

(c) and the estimated cost of such measures.

3. Write a one-page executive memo to the head of WSDA to summarize your findings.

 

Problem F (Author: Wai)

Your company has just hired by the Washington State Governor Office to study the effects of ride-sharing companies on the traffic congestion The governor wants to know if s/he wants to issue a license to a NEW ride-sharing company.

Your team is in charge of this project. Your supervisor would like you to
1. Use existing data to build a traffic congestion model for the Seattle area.
2. Investigate the effects with an additional company on the traffic congestion
3. Investigate the other possible effects with an additional company besides traffic congestion.
4. Put together a one-page memo to be sent to the governor. In the memo, explains your findings in languages that s/he can understand.