Applied Predictive Modeling

by Max Kuhn and Kjell Johnson

This is a great book on predictive modeling (i.e. statistical learning). Its strongest points are that it takes a very practical approach to the subject. This means that almost all of the discussions on algorithmic methods are accompanied with actual R code that demonstrates the execution of the discussed methods on real data. By reading and understanding this book, one will posses a good theoretical knowledge of predictive modeling with the required R skills to implement that knowledge quickly and efficiently on any dataset of the users choosing. Readers unfamiliar with this book can see what others have said here.

To learn this material as well as possible I worked through all of the book's end of chapter problems. The R code for the problems can be found free below. In addition to the R code I wrote up discussions and comments on each problem in PDF (using the mathematical typesetting language LaTeX). I converted the PDF format to a format I thought more people would find easier to read. You can preview and buy a kindle version of the book here. If you are interested in purchasing the PDF version you can do so for $41.00 (US dollars) via PayPal (see the link below). After PayPal checkout you should be taken to a web page where you can download the solution manual immediately.

• Chapter 3 (Data Pre-processing)
  
• Chapter 4 (Over-Fitting and Model Tuning)
  
• Chapter 6 (Linear Regression and Its Cousins)
  
• Chapter 7 (Nonlinear Regression Models)
  
• Chapter 8 (Regression Trees and Rule-Based Models)
  
• Chapter 12 (Discriminant Analysis and Other Linear Classification Models)
  
• Chapter 13 (Nonlinear Classification Models)
  
• Chapter 14 (Classification Trees and Rule-Based Models)
  
• Chapter 16 (Remedies for Severe Class Imbalance)
  
• Chapter 18 (Measuring Predictor Importance)
  
• Chapter 19 (An Introduction to Feature Selection)
  
• Chapter 20 (Factors That Can Affect Model Performance)