Case Study

Evaluate the use of a source indicator for resolving inconsistencies in a multi-city housing model.

Case context: You are developing a housing-price prediction model using training data collected from two different cities, Detroit and New York City. The pricing structures in these cities are highly inconsistent, meaning similar houses have wildly different prices depending on the city. You are considering adding an extra feature to each training example indicating the city of the property.

Question: Based on the source material, evaluate what effect this feature will have on the target variable y, and decide whether this is a commonly adopted solution in machine learning practice.

Sample answer: Adding the city as a source indicator feature to the input x will make the target value y unambiguous, resolving the inconsistency between Detroit and New York City data. However, you should be aware that this specific approach is not frequently used in practice.

Key points:

  • Adding a city indicator to input x makes the target value y unambiguous.
  • The feature addresses inconsistencies between Detroit and New York City data.
  • Despite its theoretical clarity, this approach is not frequently used in practice.

Rubric: The evaluation must state that the target value becomes unambiguous when the city is specified in input x, and acknowledge that this approach is not frequently used in practice.

0

1

Updated 2026-05-26

Contributors are:

Who are from:

Tags

Machine Learning

Deep Learning

Supervised Learning

Dive into Deep Learning @ D2L

Data Science

Machine Learning Strategy

Related