Overview

Model assessment, model tuning and feature engineering

Pages

Resampling	The validation set approach, cross-validation, and how much each estimate varies
Tuning	Grid, random and successive-halving search, and nested cross-validation
Data cleaning	Messy tables, missing data, useless predictors, standardization
Feature engineering	One-hot coding, interactions, splines, transforming the output
Exercises

Goals

The goal of this week is to

1. understand the validation set approach for test error estimation and hyper-parameter tuning,
2. understand the cross-validation approach for test error estimation and hyper-parameter tuning,
3. learn how to use different resampling strategies, like the validation set approach or cross-validation,
4. learn how to do hyper-parameter tuning,
5. learn how to clean data,
6. understand and learn to apply feature engineering, and
7. understand the implications of transforming the response variable.