Applying support vector machines in R comparing linear, radial, and polynomial kernels on simulated non-linear data, the Auto data set (predicting gas mileage), and the OJ data set (predicting purchase behavior), with cross-validation and error rate comparisons.

Exercises on decision trees, pruning, bagging, random forests, and BART applied to the Carseats and OJ datasets, with model comparison via cross-validation and test error.

Exercises on polynomial regression, step functions, and GAMs — applied to the Wage and College datasets, with model selection via cross-validation and formal tests for non-linearity.

Exercises on model selection and regularization comparing least squares, ridge, lasso, PCR, and PLS for predicting college applications and Boston crime rates, with model performance evaluated via cross-validation.

Resampling Methods: CV & Bootstrap (ISLR Ch. 5)

An application of classification methods (logistic regression, LDA, QDA, Naive Bayes, and KNN) to three datasets from the ISLR2 package. Includes exploratory data analysis, model fitting, confusion matrices, and comparison of test error rates across methods for predicting stock market direction (Weekly), high/low gas mileage (Auto), and high/low crime rate (Boston).

This R Markdown document contains solutions for Homework 2, covering K-Nearest Neighbors (KNN) classification vs. regression, multiple linear regression analysis on the Auto dataset, multiple regression on the Carseats dataset with qualitative predictors, and simple linear regression without an intercept. Includes diagnostic plots, model comparisons, and interpretation of coefficients.