I see many people who would like to take a glimpse at machine learning, and try to understand a bit how it works. Very often, they can either get pre-baked examples with very specific (and possibly too advanced) approaches - like deep learning - or math-oriented explanations that can be dry or just uninteresting.
I recently discovered a rather famous free textbook that I hadn't touched before: An Introduction to Statistical Learning . As you may infer by the non-glamorous title, that's a book that doesn't try to sell you something fancy about machine learning. It's quite a practical and non math-heavy introduction to most useful machine learning topics, which will lead the reader to develop an intuition for what ML methods do. SPOILER: neural networks aren't covered! So, if you're just running after the hype, that's not the book for you.
The only real drawback from the original book is that most examples and demos are coded in R. I don't especially like the language, as it is highly specialized and, most probably, you'll need to know another language beyond it for general-purpose processing.
So, I'm happy to link a couple of repositories that offer most examples from the book, but coded in Python; those should be more accessible to most people, as the language is very widespread:
There's a MOOC as well covering most of the topics from the book, by the same original authors: https://lagunita.stanford.edu/courses/HumanitiesSciences/StatLearning/Winter2016/about