Turn: Machine learning materials Books

Links: Http://suanfazu.com/t/topic/15 Starter Book list

2. The beauty of mathematics PDF586
   The author Wu Everyone is familiar with it. The application of mathematics in the fields of machine learning and natural language processing is described in a very popular language.

4. "Programming Collective Intelligence" ("collective Intelligence Programming") PDF343
   Author Toby Segaran is also the author of Beautifuldata:the Stories Behind Elegant Data Solutions (the beauty of data: The story behind the decryption of elegant data solutions). The greatest advantage of this book is that there is no theoretical derivation and complex mathematical formulas, it is a very good introductory book. At present, the Chinese version has been out of stock, for those who are interested in this field, the English PDF is a good choice, because there are a lot of classic book translation is poor, can only read English version, it is better to start from this. Also, this book is suitable for quick reading, because it is commented that after reading some classical books with mathematical deduction, the book will find nothing to say, just give a lot of examples.

6. Algorithms of the Intelligent Web (Smart Web algorithm) PDF138
   Author Haralambos Marmanis, Dmitry Babenko. The formula in this book is a little bit more than "collective intelligence programming", the example of which is mostly the application on the Internet, to see the name. The disadvantage is that the matching code inside is BeanShell and not python or anything else. In general, this book is still suitable for beginners, and the same need to read the same as the previous one, if you finish reading the previous one, this one can not have to look at the code, understand the main idea of the algorithm is OK.

8. "Statistical learning Method" PDF Blur 281 epub178
   Author Hangyuan Li, is one of several people in the field of machine learning in China, has been a senior researcher at MSRA, now Huawei Noah's Ark laboratory. The book wrote 10 algorithms, each of the introduction of the algorithm is very straightforward, directly on the formula, is downright "dry book." The references at the end of each chapter are also convenient for children's shoes that want to understand the algorithm directly to the classic paper; This book can be used as a supplementary reading for each of the two books.

10. "Machine learning" (ml) PDF520
    Author Tom Mitchell is a master of CMU, with a machine learning and semi-supervised learning Network course video. This book is a good book for translation in the field, and the algorithms are much larger than the range of statistical learning methods. It is commented that the book is mainly about inspiration, explain why the formula was founded rather than derivation; But some basic classics are still not outdated, so this book is now almost a must-read for machine learning.

12. "Mining of Massive Datasets" ("Big Data") PDF172
    Author Anand rajaraman[3], Jeffrey David Ullman,anand is the PhD of Stanford. This book introduces a lot of algorithms, and also describes these algorithms in the data size is larger than the time of the deformation. But in space, each algorithm does not have a sense of unfolding, if you want to know more about the need to look at other information, but then the algorithm is enough to understand. There is also a little bit of a lack of the original and translation of the book have many errors, errata longer, the reader must attentively.

14. Mining:practical machine learning Tools and techniques (Data mining: Utility learning Technology) PDF153
    Author Ian H. Witten, Eibe Frank is the author of Weka and a professor at the University of Waikato in New Zealand. Their "managinggigabytes" [4] is also a classic book on information retrieval. The most characteristic of this book is the introduction of the use of Weka, but its theoretical part is too thin, as an introductory book can also, but the classic introductory books such as "collective Wisdom Programming", "Intelligent Web Algorithm" has been very classic, learning words should not read too many introductory books, suggest only to see some of the above two books did not talk about algorithms.

16. "Machine learning and its application"
    Zhou Zhihua, Prof Qiang Yang editor. From the "machine learning and its Application seminar" Anthology. The seminar was initiated by the Fudan University's Intelligent Information Processing laboratory, which has been held for 10 sessions, and domestic Daniel such as Hangyuan Li, Xiangliang, Wang Haifeng, tie and Kaiyu have lectured at the conference. This book speaks of a lot of machine learning at the forefront of specific applications, need to have a basic ability to understand. If you want to learn about machine learning trends, you can browse the book. Academic conferences in the area of interest are the way to discover research trends.

18. "Managing Gigabytes" (Deep search engine) PDF
    A good book for information retrieval.

20. "Modern Information Retrieval" PDF
    Ricardo Baeza-yates et al. 1999. It looks like the first book that covers IR completely. Unfortunately, IR has progressed rapidly over the years, and the book is somewhat outdated. It's a good idea to turn over and do a reference. In addition, Ricardo is now the head of Yahoo! for Europe and Latin Ameria.

22. Recommended Systems Practice PDF164
    Xiangliang, good introductory reading.

Depth

2. Pattern Classification (second edition of Pattern classification) PDF118
   Author Richard O Duda[5], Peter E. Hart, David. The foundation of pattern recognition, but the better method of SVM and boosting method is not introduced in the recent dominant position, and is evaluated as "exhaustive suspicion".

4. "Pattern Recognition and machine learning" PDF
   Author Christopher M. Bishop[6], abbreviated to PRML, focuses on probabilistic models, is a Bayesian method of the tripod, according to the evaluation "with a strong engineering breath, can cooperate with Stanford University Andrew Ng's machine learning Video tutorial to learn together, the effect of doubling. ”

6. The Elements of statistical learning:data Mining, Inference, andprediction, (Statistical Learning Fundamentals: Data Mining, reasoning and Forecasting, second edition) PDF
   Author Roberttibshirani, Trevor Hastie, Jerome Friedman. The author of this book is the most active researcher of the boosting method, and the invention of gradient boosting presents a new perspective of understanding boosting method, greatly extending the scope of application of the boosting method. This book is a more comprehensive introduction to the most popular methods, and it may be a little more useful to the engineering staff. On the other hand, it not only summarizes some of the technologies that have matured, but also has a brief discussion on some of the issues that are still being developed. Let the reader fully realize that machine learning is a still very active field of research, it should make academic researchers often read often new feelings. "[7]

8. "Data mining:concepts Andtechniques" (the third edition of Mining: Concepts and Technologies) PDF
   Author (United States) Jiawei Han[8], (plus) Micheline Kamber, (plus) Jian Pei, of which the first author is Chinese. The book is no doubt the data mining aspects of the classic, but the translation is always sprayed, no way, most of the translated books are sprayed, want to eat other people chew things, learn English well.

10. "AI, Modern Approach 2nd" PDF
    Peter Norvig, a undisputed field classic.

12. "Foundations of statistical Natural Language processing" PDF
    The field of natural language processing is a recognized classic.

14. "Information theory:inference and Learning algorithms" PDF

16. "Statistical learning theory" PDF
    Vapnik's masterpiece, the authority of the statistical academia, this book to the theory to the philosophical level, his other book "The Nature Ofstatistical Learning theory" is also a rare statistical study of good books, but these two books are relatively deep, Suitable for readers with a certain foundation.

Fundamentals of Mathematics

2. Matrix Analysis PDF246
   Roger Horn. The undisputed classical matrix analysis field

4. "Probability theory and its application" PDF
   William Feller. A very good book, but the math is too heavy for machine learning.

6. "All of Statistics" PDF scanned version
   HD version 164
   In this direction of machine learning, statistics are also very important. Recommend all of statistics, a very concise textbook for CMU, focusing on concepts, simplifying calculations, simplifying concepts and statistical content unrelated to machine learning, which can be said to be a good quick start material.

8. "Nonlinear programming, 2nd" PDF
   Optimization method, a reference book for nonlinear programming.

10. "Convex optimization" PDF support code
    Boyd's classic books, cited more than 14,000 times, for practical applications, and have matching code, is a rare good book.

12. "Numerical optimization" PDF
    The second edition, Nocedal, is ideal for students and engineers of non-numeric majors, with clear and detailed algorithm flow and clear principles.

14. "Introduction to Mathematical Statistics" PDF
    Sixth edition, Hogg, this book introduces the basic concepts of probabilistic statistics and various distributions, as well as Ml,bayesian methods.

16. "An Introduction to probabilistic graphical Models" PDF118
    Jordan, this book introduces the basic concepts of conditional independence, decomposition, blending and conditional blending, and also introduces the implicit variables (latent variables), and I believe that we have encountered this concept when we implement EM algorithm in the hidden Markov chain and the Gaussian mixed model.

18. "Probabilistic graphical Models-principles and techniques" PDF
    Koller, a very thick and comprehensive book, is highly theoretical and can be used as a reference book.

20. Specific Math PDF
    Classic

Turn: Machine learning materials Books