Theoretical Machine Learning (Fall 2026)

Quiz Overview

• There will be several short, closed-book, individual-effort, in-class quizzes throughout the semester. The timing of the quizzes will not be announced in advance, in part to encourage regular lecture attendance. Together, they account for 30% of your final grade.

• The purpose of the quizzes is to assess your understanding of the homework. Quizzes will take place after the corresponding homework is due and its solutions have been released.

• Quiz questions do not simply repeat homework problems. Instead, they will ask short questions related to the ideas, techniques, and solutions from the homework. They are designed to be substantially easier than the homework problems themselves and should take about 10–15 minutes to complete. Examples include:

  • Explain the role of a particular step or assumption in a homework solution.

  • Fill in a missing step in an argument from a homework solution.

  • Apply an idea or technique from the homework to a simple variant of the original problem.

  • Determine whether a small modification of a statement from the homework remains true, and briefly justify your answer.

• Any formulas or definitions needed for the quiz will be provided; the quizzes are intended to test your understanding, not your ability to memorize formulas or definitions.

• Your lowest quiz score will be dropped. This includes a score of zero resulting from a missed quiz.