Video: Bias-Variance trade-offΒΆ

Note: Subtitles are available (click CC button at the bottom of the video)

Note

The material in this video is based on Chapter 2.2 of the ISLR book (cited in the teaching materials for this course). I encourage you to read Chapter 2 to learn more details about the bias-variance trade-off.

After watching the video lecture, please, try answering the following questions:

When a more flexible model will be preferred over a less flexible model?

  • the shape of Conditional Expectation Function in the population is linear in parameters

  • the shape of Conditional Expectation Function in the population is highly non-linear in parameters

  • the variance of noise around the Conditional Expectation Function in the population is high

  • the variance of noise is low

  • the number of observations per number of parameters (\(n/p\)) is high

  • (\(n/p\)) is low