Learn Before
Case Study

Choosing the right split for model selection decisions.

Case context: A hospital analytics group is developing a model to predict whether patients will return within 30 days. They already have a labeled dataset and want to decide which input variables to keep and which model settings to adjust.

Question: Which dataset split should they use to compare modeling choices and make those selections, and what is that split also called?

Sample answer: They should evaluate options on the development set, which is also known as the validation set or hold-out cross-validation set, because it is used for choosing features and tuning model settings.

Key points:

  • Names the development set as the correct split.
  • Gives another acceptable name: hold-out cross-validation set.
  • Explains that feature selection and parameter tuning are performed using this split.

Rubric: The answer should identify the development set and state its alternative name as the appropriate split for feature selection and parameter tuning.

0

1

Updated 2026-08-12

Contributors are:

Who are from:

Tags

D2L

Dive into Deep Learning @ D2L

Machine Learning

Deep Learning

Supervised Learning

Data Science

Machine Learning Strategy

Machine Learning Yearning @ DeepLearning.AI