Introducing Lit-LLaMA: a minimal, optimized rewrite of LLaMA licensed under Apache 2.0 →

Log in or create a free account to track your progress and access additional course materials  

Unit 3 Exercises

Exercise 1: Banknote Authentication

In this exercise, we are applying logistic regression to a banknote authentication dataset to distinguish between genuine and forged bank notes.

The dataset consists of 1372 examples and 4 features for binary classification. The features are

  1. variance of a wavelet-transformed image (continuous)
  2. skewness of a wavelet-transformed image (continuous)
  3. kurtosis of a wavelet-transformed image (continuous)
  4. entropy of the image (continuous)

(You can fine more details about this dataset at

In essence, these four features represent features that were manually extracted from image data. Note that you do not need the details of these features for this exercise.

However, you are encouraged to explore the dataset further, e.g., by plotting the features, looking at the value ranges, and so forth. (We will skip these steps for brevity in this exercise) Link to exercise notebook: exercise_1_banknotes.ipynb

Exercise 2: Standardization

This exercise is an extension of Exercise 1. Here, the goal is to add code to standardize the features such that they have a mean of 0 and a standard deviation of 1 as discussed in Unit 3.7.

Link to exercise notebook: exercise_2_standardization.ipynb

Log in or create a free account to access:

  • Quizzes
  • Completion badges
  • Progress tracking
  • Additional downloadable content
  • Additional AI education resources
  • Notifications when new units are released
  • Free cloud computing credits

Unit 3 Exercises

Questions or Feedback?

Join the Discussion