Lecture 25 – Inference for Modeling

Presented by Fernando Perez and Suraj Rampure

Content by Suraj Rampure, Fernando Perez, John DeNero, Sam Lau, Ani Adhikari, Deb Nolan

A reminder – the right column of the table below contains Quick Checks. These are not required but suggested to help you check your understanding.

Video Quick Check
25.1
A big picture overview of inference. Parameters and estimators. Bias and variance of estimators. The sample mean estimator.
25.1
25.2
Using bootstrap resampling in order to estimate the sampling distribution of an estimator.
25.2
25.3
Defining confidence intervals more generally. Describing and demoing how we can use the bootstrap to create confidence intervals for population parameters.
25.3
25.4
The assumptions we make when modeling with linear regression..
25.4
25.5
Using the bootstrap to estimate the sampling distributions of parameters in a linear regression model. Inference for the true slope of a feature.
25.5
25.6
Multicollinearity, and its impacts on the interpretability of the parameters of our model. A summary of the lecture, and a brief overview of the ML taxonomy.
25.6