Without a doubt my beloved Python is the Lingua Franca for machine learning these days. Tons of create tools by Google, Facebook and other big players make it easy to learn about machine learning. All offer a great Python API. At Amazon you find 200+ books teaching you stuff about machine learning using Python. A lot of material to choose from. And from my point of view, also a lot of material to fail with. Let me recommend two books I find particularly helpful, when you want to learn something about these fascinating fields and start to dive into one of the hottest topics in tech at the moment.
Deep Learning with Python
This is my prefered recommendation because it is strictly focused on Deep Learning, which is actually the “hot shit” in the AI field at the moment. Everyone is doing it, everyone is telling some truths and some myths about it.
And that is where the book of François really shines. He uses the hot tools Tensorflow and Keras. (He is the creator of Keras.) François works for Google where a lot of deep learning is utilised and researched currently. So he knows what is possible and what is not.
I really love how he gives a good overview about deep learning in the first part of his book. Clarifying what AI is, what machine learning is and what deep learnings. How do these buzzwords relate to each other. He clarifies some of the myths currently pushed on the internet by semi knowledgable people. You get a good technical introduction into machine learning, too. And you don’t need a maths degree to follow the book. After you have read part one of the book, you have a good grasp what you are talking about.
The second part describes some of the common use cases for deep learning. You learn about using this technology in computer vision, text processing and in generating contents likes images and text.
Of course all the examples use Keras, but this is also the best choice from my point of view to interact with Tensorflow. Tensorflow itself has a great Python API. But the API of Keras is awesome, easy to use and so pythonic.
Introduction to Machine Learning with Python
You might ask why I recommend another book after my hymn of praise about the book of François Chollet. Well, the first book completely focuses on deep learning which is a subset of machine learning. This book gives a much wider introduction into the field of machine learning. It also goes into more detail about various aspects – different training methods, data representation, different ways of modelling, neural networks, etc.
I also like, that this book that utilises another great machine learning library – scikit-learn. It has its origins in the great scientific toolchain for Python which has been available for years. So besides the actual machine learning toolkit, you also learn a bit about NumPy, Pandas, Jupyter notebooks etc.
So, if you prefer a more technical and wider approach to get started with machine learning, this book is definitely for you.
I think I have bought and read almost all books about Python and machine or deep learning. But to get started I will still recommend the two books above. There are some nice cookbooks around or even some cheaper books. But these two are great to learn and the authors know how to teach a complicated topic.
Finally, I just want to mention one book, that I particularly enjoyed – Deep Learning for Computer Vision with Python. If you like to learn about utilising deep learning and Python for anything in the field of Computer Vision. Read this book! But read an introductory book first. 😉
Image credit: Dushan Hanuska. Shared under the Creative Commons Attribution-ShareAlike 2.0 Generic license.