Books on nonparametric statistics

A broad view of nonparametric methods:
1. All of Nonparametric Statistics by Larry Wasserman
  - As along as you have background in mathematical statistics (upper level undergraduate), this book should be easy to follow.
2. All of Statistics by Larry Wasserman
  - This book also covers part of the nonparametric approach; this book is featured with an elegent overview on most modern statistical methods (e.g. graphical model).
More specialized to density estimation (and kernel density estimator):
1. Multivariate Density Estimation: theory, practice, and visualization by David Scott
  - One of the best book for density estimation; definitely worth reading it if you are interested in nonparametric statistics.
More about nonparametric theory:
1. Introduction to Nonparametric Estimation by A.B. Tsybakov
  - This book is a bible for nonparametric methods and theories; a great book for PhD students.
2. Asymptotic Statistics by van der Vaart
  - This is a bible for asymptotic theory (also covers some nonparametric theory); I would recommend reading this book after taking all the statistics PhD theory courses.
Specialized theory books:
- I would not recommend to read through any of the following books unless you have lots of time. These books are excellent references but they might be too technical.
1. Combinatorial Methods in Density Estimation by Luc Devroye and Gabor Lugosi
  - This book dedicates more on “concentration inequality”-type method for density estimation theory; I would recommend to read Book #4 first.
2. A probabilistic Theory of Pattern Recognition by Luc Devroye, Laszlo Gyorfi, and Gabor Lugosi
  - This book has many theories about k-nearest neighbor (k-NN) method. If you want to learn theory about k-NN, this will be the right reference.
3. A Distribution-Free Theory of Nonparametric Regression by Laszlo Gyorfi, Michael Kohler, Adam Krzyzak, and Harro Walk
  - This book focuses on theory of nonparametric regression. It has many useful contents on least square estimator (and penalized version). If you work on nonparametric regression, this will be a great reference.