Is this list still valid to date?
Hello Everyone!
Okay, I have read all the books Paul mentioned above and here is my feedback:
Deep Learning by Ian Goodfellow et al.: If you are new to machine learning, this book is not for you. Even if you are a machine learning practitioner (the one who uses current techniques and doesn’t research), I won’t suggest reading this book. This book is for researchers who are already familiar with ML techniques and willing to advance.
Deep Learning with Python by François Chollet: Yep, I highly recommend this book to newbies. It has a GitHub repo as well. But this book does not cover maths, so, I suggest first taking Deep Learning Specialization and then reading this book.
The Mechanics of Machine Learning by Terence Parr and Jeremy Howard: This is a good book but it is still incomplete, as Paul mentioned, “book in preparation”. And, I think the authors are too busy to finish it as it has been five years. So, you can skip this book.
Neural Networks and Deep Learning by Michael Nielsen: Another solid choice, but I recommend reading it after completing the Deep Learning Specialization.
Hands-On Machine Learning with Scikit-Learn, Keras and Tensorflow by Aurélien Géron: Absolutely excellent and has a GitHub repo too.
Alright. If you ask me to suggest only one book, I will say go for Aurélien Géron’s book (latest edition). It covers a lot of things of TensorFlow Developer Professional Certificate, TensorFlow: Advanced Techniques Specialization, TensorFlow: Data and Deployment Specialization, Natural Language Processing Specialization, Generative Adversarial Networks (GANs) Specialization, and of course Machine Learning Specialization, and Deep Learning Specialization. It also covers auto-tuning. So, it is almost an all-in-one guide.
Still, you may need to take DLS to understand some maths. You can do both simultaneously (reading the book in the morning and taking the course in the evening). It has approx. 700 pages (third edition) and if you read 10 pages daily, and practice the code from Géron’s GitHub (very important), you can complete this book in 70 days. Go for it! But don’t forget to play with his code—working with the code is essential. Without it, you risk wasting your time.
Best of luck.
Best,
Saif.
6 Likes
Amazing content you put in there. Respect.
Btw I’ve completed Machine Learning Specialization. Concepts of maths were discussed surely, but deep level discussion related to calculus or statistics is missing. So when it comes to maths I would definitely attend some more advanced course such as: Mathematics for Machine Learning and Data Science Specialization, or anyone has some better suggestions?
@icybergenome I am doing this: Mathematics for Machine Learning and Data Science | Coursera
He is pretty good, no nonsense.
1 Like
I hesitate to recommend that course, because it is very nice mathematics but a significant portion doesn’t apply to either machine learning or data science.
Note also that the notational conventions used in that course are not consistent with other DLAI course machine learning practices. This can cause confusion.
1 Like
Adding to this list a newsletter about advanced deep learning topics. They do not post very often, maybe an article every 4-5 months but their articles are really interesting and worth reading.
Best,
Rosa
1 Like