<< 1 >>
Rating: 
Summary: excellent text on density estimation
Review: I had the good fortune to take a short course from Bernie Silverman on density estimation just after this book came out in 1986. It is one of the clearest treatments of the subject and I found it particularly good on the coverage of optimal kernels. It is also filled with good practical examples and advice. For instance, the Old Faithful data provides an excellent example of a bimodal distribution where kernel density estimation provides a way to detect the two modes.The author was also very perceptive in recognizing the value of projection pursuit techniques and bootstrap methods and the way density estimation techniques relate to these methods. The book has the virtue of being clear and concise.
Rating:
Summary: excellent text on density estimation
Review: I had the good fortune to take a short course from Bernie Silverman on density estimation just after this book came out in 1986. It is one of the clearest treatments of the subject and I found it particularly good on the coverage of optimal kernels. It is also filled with good practical examples and advice. For instance, the Old Faithful data provides an excellent example of a bimodal distribution where kernel density estimation provides a way to detect the two modes. The author was also very perceptive in recognizing the value of projection pursuit techniques and bootstrap methods and the way density estimation techniques relate to these methods. The book has the virtue of being clear and concise.
Rating:
Summary: Best book on this subject
Review: Quite a few books have been written since 1986, but this book is still the best. Very intuitive and very readable. It is written with a mastery of the subject and an excellent style of pedagogy. I remember of the joy and refreshness of reading this book around 1987 and it has served me well on a very important introductory of mordern statistics without having to go through tedious "math" notations and a shining example that statistics can be full of intuitive ideas and beautiful. For people unfamiliar with this book, it deals with probability density estimation using the idea of "local averages", and so it does not deal with other techniques such as splines. Also it is purely a density estimation book, and does not deal with another important problem, namely regression estimation (on which there are many other books). In summary, this book introduces the ideas and sense of "smoothing", a large (perhaps a little overblown) area of modern statistics. If you want to learn statistical smoothing, besides from Steve Marron, this one is the way to go.
Rating:
Summary: Best book on this subject
Review: Quite a few books have been written since 1986, but this book is still the best. Very intuitive and very readable. It is written with a mastery of the subject and an excellent style of pedagogy. I remember of the joy and refreshness of reading this book around 1987 and it has served me well on a very important introductory of mordern statistics without having to go through tedious "math" notations and a shining example that statistics can be full of intuitive ideas and beautiful. For people unfamiliar with this book, it deals with probability density estimation using the idea of "local averages", and so it does not deal with other techniques such as splines. Also it is purely a density estimation book, and does not deal with another important problem, namely regression estimation (on which there are many other books). In summary, this book introduces the ideas and sense of "smoothing", a large (perhaps a little overblown) area of modern statistics. If you want to learn statistical smoothing, besides from Steve Marron, this one is the way to go.
<< 1 >>
| 
