2018.09.17-21 ISL Chapter 2 Plan

$\DeclareMathOperator{\Var}{Var}$ $\DeclareMathOperator{\Cov}{Cov}$ $\DeclareMathOperator{\Span}{Span}$
  • Download ISL book.
  • Start reading Chapter 2. Lectures will give you a basis, but the depth of understanding comes from reading.
  • Give feedback on pace of the chapter.

Weekly Reading

  • Introduction, pages 8-9 (pdf 22-23): premises and who should read the book.
  • 2.1 What is Statistical Learning? Some material may be familiar from the video of Andrew Ng.

    • What is the difference between parametric and nonparametric models?
    • What is the tradeoff between flexibility and interpretability?
    • Supervised vs. Unsupervised.
    • Regression vs Classification.
  • 2.2.1: Assessing Model Accuracy (important): Mean Squared Error (MSE). Graphs showing train vs test MSE.

  • 2.2.2: Bias-Variance Tradeoff

  • 2.2.3: Classification. Discuss in class. K-nearest neighbors. Bayes decision boundary(?)

Weekly Homework (due next Monday)

  • Work through Lab 2.3
  • Write your own “study guide” for Lab 2.3 as discussed in class and below.
  • Chapter 2 Problems 1-10. (Do number 8 immediately after finishing Lab 2.3.)

Chapter 2 Study Guide

Write a study guide, probably with a partner. The study guide should show how to make variations on the ideas from section 2.3 of the book, so you become comfortable with it.

Examples:

  1. Start with a square matrix full of random numbers 10-100. (2.3.1) Choose them uniformly (all equally likely).
  2. Find the vector created by adding the first column and the second row. (2.3.3)
  3. Write a function that returns the value in the square one row and two columns before the last (lower right) element in the array. (2.3.3)
  4. Make a mesh grid of all possible pairs of x = [1,5,10] and y=[2,20,40]. (2.3.2)
  5. Amusement: Look up a cool function for 3D graphing. One to try is $$f(x,y) = \frac{-1}{x^2+y^2}$$