Machine Learning for Algorithmic Trading: Predictive models to extract signals from market and alternative data for systematic trading strategies with Python

As someone who's been interested in both machine learning and financial markets, I was looking for a comprehensive resource that would bridge these two worlds. Stefan Jansen's "Machine Learning for Algorithmic Trading" not only delivered on this promise but exceeded my expectations by enabling me to build my own profitable algorithmic trading bot.What sets this book apart is how it provides a complete end-to-end workflow. The progression from basic concepts to advanced implementation is logical and thorough. Jansen starts with essential market data handling, moves through feature engineering techniques, and culminates in sophisticated model development and backtesting. The Python code examples using popular libraries like pandas, scikit-learn, and PyTorch provided immediate practical value rather than just theoretical concepts.I particularly appreciated the diverse range of ML techniques covered - from traditional algorithms to deep learning approaches. The sections on feature engineering and alpha factor research were especially valuable for my trading bot development. The book doesn't just teach you algorithms; it shows you how to apply them meaningfully to extract signals from market data.The inclusion of backtrader and Zipline for strategy testing was instrumental in helping me validate my ideas before risking real capital. I was able to iterate on my strategies, identify weaknesses, and refine my approach using the framework provided in the book.While the book is certainly dense at 800+ pages, it serves as both a learning resource and a reference manual. I continue to revisit specific chapters as I enhance my trading strategies. Even with some prior knowledge in both programming and finance, I found tremendous value in Jansen's comprehensive approach.One small caveat: some of the code examples require updating as libraries evolve, but the core concepts remain solid and adaptable. In fact, working through these updates enhanced my understanding of the underlying systems.Bottom line - this book delivered exactly what I needed: the knowledge and tools to transform my interest in ML and markets into a functioning algorithmic trading system. For anyone serious about applying machine learning to trading, this book is an essential investment that can potentially pay for itself many times over.

What I love about this book is that it delves into light math theory before diving into the Python code. This book is good for someone with a intermediate Python background and machine learning or finance knowledge. From there, this book will help fill in the gaps. I highly recommend this book!

As with all code related books, this one is started to show its age. Some of the links are no longer available and some of the code does not work. If you know what you are doing, you should be able to make it work and keep going, but finding the exact same data sources might be a little more difficult. I'm only up to the second chapter so far, since the first thing I did was to create my docker environment and pull all of the data sources.

I enjoyed this book b/c it was very informative. It helped me to understand machine learning better and when and where it could be useful. I recommend this book to anyone who wants to learn more about machine learning and how to apply it.

For example, yes all of the photos are black and white. However in the preface there's very clear instruction on where you can find the color versions (PDFs in the github repo, for those who eschew prefaces) and if you intend to use any of the Python code that goes along with this tome, you'll see the color versions can often also be found in the Jupyter notebooks - a fact frequently referenced in the first two chapters.Second, getting python environments up and running smoothly is, unfortunately, rarely a very easy task. This is certainly not exclusive to this particular use case.If I had any complaint at all about the book, it's that it is overly thorough so you may find yourself slogging through some tedium as you begin. It's not broken up in a way that easily allows for skipping ahead (at least not for my prior knowledge set). In the first couple chapters I've found I've needed - on average - every other paragraph and that the subject matter is the source of the dryness, not the author's use of language; which so far has been smooth and flows far more gracefully than my own.I will update after more extensive use of the author's code.

The book overall is very didactic, my only recommendation would be to use a more simple set up as many of the recommended tools and libraries are outdated, not the author’s fault but renders impossible to follow some of the examples. A more simple set of standard libraries, perhaps could be more stable and allow to follow better the presented examples.

Haven't fully read, but it's great that the author offers a free pdf through the publisher and through his personal git page. Hope to apply this material soon.

Content is great yet book printed in grayscale making all plots/charts unreadable.