Context

HomeRate uses a regression model to predict apartment rental prices so leasing agents can set listing prices faster. Two candidate models are performing similarly overall, but the team is unsure whether to optimize for RMSE or MAE because a small number of large pricing mistakes create business risk.

Current Performance

The Problem

Model A has lower average absolute error for typical listings, but Model B has fewer very large misses. The pricing team wants a recommendation on which metric should drive model selection and monitoring.

Requirements

1. Explain the difference between MAE and RMSE in practical terms.
2. Interpret the metric tradeoff between Model A and Model B using the table above.
3. Recommend when MAE should be preferred and when RMSE should be preferred.
4. State which model you would choose for HomeRate and justify the decision.
5. Suggest how you would validate that the chosen metric aligns with business impact.

Constraints

• Overpricing a unit by more than $400 can delay occupancy and reduce monthly revenue.
• Underpricing by more than $400 reduces realized rent and is hard to recover.
• The operations team can tolerate small routine errors, but large outliers are costly.