The best book about machine learning I’ve found so far is ‘Hands-On Machine Learning with Scikit-Learn, Keras, and Tensorflow’.
This is a complete manual that treats supervised learning, unsupervised learning and deep learning. It is divided into 19 chapters, each consisting of a particular model or technique you can use in your projects. It really covers a wide range of topics: from simple linear regression with polynomial features to decision trees, from logistic regressions to convolutional neural networks.
What makes this book so good is that not only it shows you the technique and how to deploy it with the renowned packages already available in Python, but also it provides you with an equivalent code that does the same thing, so that you can understand what’s going on behind the scenes. Also, it explains the mathematical algorithms and operations that lie behind the simple deployment of a model. Moreover, you can find the jupyter notebooks with all the code online, so the results in that book are reproducible.
I strongly recommend that you read that book in front of your laptop, so you can copy the code provided step by step and understand it deeply. In fact, if on one hand it’s a complete book, on the other it’s very complex to follow and, if you really want to master the topics it deals with, you have to be prepared to spend a lot of time reading it.
To sum up, I recommend this book to all machine learning and AI enthusiasts like me, who would like to improve their skills and gain an insight into such marvelous topics.