Learning Machine Learning, ML Books & Codes
MLAPP (Kevin Murphy), Machine Learning - A Probablistic Perspective, is more comprehensive, insightful and interesting, and contains more "real" examples/problems. However, the presents are kinda out of order, which can be difficult to follow for a first book.
ESL (Trevor Hastie), Elements of Statistical Learning, 2n ed. Springer 2017. Contains even harder math (ref: zhihu ans. --> point.2), stats.-oriented, advanced.
Good Text Books (for Master & PhD ?)
- Machine Learning by Peter Flach (Cambridge University Press 2012) - The Art and Science of Algorithms that Make Sense of Data. [The examples in this book are clear and easy to understand.]
- Machine Learning by Tom M Mitchell (McGraw-Hill 1997). (Classical, many original terms and ideas.)
- PRML by Christopher Bishop. (See "favoured")
南京大学 周志华 《机器学习》 (“西瓜书”) 清华大学出版社 2016：有伪码，但是比 Peter Flach 的难懂，需要比较熟悉统计语言，旁注格式不如他的另一本书《集成学习》中的 in-line 注释。 类似清华的红皮 PR，没有推导过程，推导可以参考开源南瓜书：西瓜书的公式推导 (“南瓜书”) by Datawhale。
- madhug-nadig/Machine-Learning-Algorithms-from-Scratch 2018
- kaggle: Machine Learning Algorithms from Scratch 2020
- Yoshua Bengio 和 GAN 之父 Ian Goodfellow 等人合著的 Deep Learning: 中文版 on github (pan.baidu), 英文版 github (pan.baidu).
- DL With Python, by Francois Chollet (also the author of Keras). Detailed introduction of CNN & RNN on CV & text analysis in Keras.
- DL for CV with Python [//trilogy], by Adrian Rosebrock.
- Hands on ML with SL and TF [//genius for clarity], by Aurélien Géron.
- DL - A Practioner's Approach, [java], by Adam Gibson & Josh Patterson.
- DL with Python, by Jason Brownlee. [//with code, easy to understand]
- Practical Python and Open CV, by Adrian Rosebrock. [// fast into CV]
- DL with TF, by Giancarlo Zaccone, [// DL 101].
- Pattern Classification by Richard O. Duda, David G. Stork, Peter E.Hart, 2015.
- A Course in Machine Learning by Hal Daumé III, 2017.
- Programming Collective Intelligence 《集体智慧编程》 2007 O'REILLY by Toby Segaran: Programming selected algorithms from scratch. Good to read when learning some ML courses (e.g. ML by Andrew Ng, Stanford).