Potentially useful exercises from Bishop's book

[sandylaker]

Here are some curated exercises from Bishop's book. They might be insightful for the SL lecture.

• Chapter 2

  • 2.21 Good exercise for entropy
  • 2.22 Constrained optimization application in maximization of entropy. Quite classic exercise.
  • 2.23-2.25 Good combo exercise for entropy.
  • 2.26 Good exercise for KL-Divergence.
  • 2.41 Regularized linear regression from a Bayesian perspective.

• Chatper 4:

  • 4.3 Exercise for logistic regression.
  • 4.5 Exercise for linear regression.
  • 4.7 Multi-output linear regression.
  • 4.8 Exercise for multi-output regression.
  • 4.11 Exercise for MLE vs. linear regression.

• Chapter 5

  • 5.7 Risk minization for classification.
  • 5.10 Risk minization for classification.
  • 5.13-5.15 MLE and Naive Bayes.
  • 5.16-5.17 Good combo exercise for MLE.
  • 5.20 Exercise for Logistic regression.

• Chapter 9

  • 9.2 Relation between weight decay and ridge regression.
  • 9.6 Maybe useful for second-order optimization technique in optimization course.
  • 9.18 Linear regression with dropout regularization.