Characterizing Colleges

Author

Rubric

Published

March 10, 2025

Multiple Regression

  • The student:
    • Minimally computes the RMSE of more than one model
    • Discusses how to RMSE relates to relates to the data set in human readable language
    • Discusses one claim that may be made about the data based on these multiple RMSEs.
    • Supports any numerical claims made with code.

Feature Engineering

  • The student:
    • Minimally adds 6 new features.
    • Removes all non-engineered features.
    • Calculates an RMSE and compares it to the earlier RMSEs.
    • Comments on independence of features or achieves independence through engineering.

Naive Classification

  • The student:
    • Minimally provides \(K\)-NN and/or Naive Bayes specialized features.
    • Reports Kappa values for more than one method.
    • Provides a narrative for the difference in Kappa values.

Improved Classification

  • The student:
    • Calculates relative frequencies of some categories.
    • Derives weights from these categories.
    • Trains and new model with novel Kappa value.
    • Provides a narrative for the difference in Kappa values.

Ethics

  • The student:
    • Writes at a college level.
    • Does not contradict code samples.