Statistical Learning: 5.Py Bootstrap I 2023  Summary and Q&A
TL;DR
Bootstrap is a general procedure used to measure the variance of complicated statistics by sampling data with replacement.
Questions & Answers
Q: What is the purpose of using the bootstrap in measuring variance?
The bootstrap is used when there are no closed form expressions to calculate the variance of a complicated statistic. It allows for estimating the variance by sampling the data with replacement and computing the statistic for each sample.
Q: How does the bootstrap work in practice?
The bootstrap involves randomly sampling the rows of a data frame with replacement many times. For each sample, the statistic of interest is computed, and the sample standard deviation of these estimates is used as an estimate of the variance.
Q: What does the "boot SE" function do?
The "boot SE" function is provided as an example to automate the bootstrap process. It takes an estimator function, a data set, and the number of samples as inputs. It applies the estimator to different samples and collects the results to compute the standard error.
Q: How can the bootstrap be used for linear regression?
The bootstrap can also be used for linear regression by fitting a regression model to a data frame and a set of indices. The "boot SE" function can be used to compute the standard error in this case as well.
Summary & Key Takeaways

The bootstrap is used to measure the variance of a statistic, particularly in an investment problem where two assets are being compared for optimal investment based on variance.

To use the bootstrap, random samples with replacement are taken from the data and the statistic is computed for each sample. The sample standard deviation is then used as an estimate of the variance.

A function called "boot SE" is provided to automate the bootstrap process and compute the standard error of the statistic.