Elden, L (2007). Matrix Methods in Data Mining and Pattern Recognition. Philadelphia (USA): Society for Industrial and Applied Mathematics
暂无分享,去创建一个
Matrix Methods in Data Mining is a quite recent book and a short but interesting read for those interested in the application of modern matrix methods in pattern recognition and data mining problems. It is stated in the book that this was written primarily for undergraduate students who have previously taken an introductory scientific computing/numerical analysis course and it is surely a good place for such students to look for final year projects ideas or to start research in the subject. Also for those students and lecturers that are interested, the author provides a website where a collection of exercises and computer assignments are available. The book is meant not to be primarily a textbook in numerical linear algebra but rather an application-oriented introduction to some techniques in modern linear algebra with the emphasis on data mining and pattern recognition. It should be noted also that the book is of very limited size, around 220 pages, so the reader should not expect a full account of the mathematical and numerical aspects of the algorithms used, but an introduction and its use for real life problems. A very valid thing to point out is that the book provides MATLAB scripts demonstrating how to implement certain algorithms, but it should not be seen as a book of recipes. Its aim is to provide students with a set of tools that may be tried as they are but most likely will need to be modified to be useful for a particular application. That is specially true if the reader intends to develop a solution that runs as fast as possible in MATLAB as the scripts were clearly designed to focus on cleanliness being then as easy to