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

Initially - I thought it would be a book that would touch the basics only. I was very pleasantly surprised. Fantastic read and highly recommended high quality content

Amazing book, detailed and has everything I needed and more

As one US reviewer stated quite rightly "For anyone that wants this book you should understand that this book is what you make of it.", there is NO hand-holding in this book, you are expected to get on with it and read-around as required/needed, it may be reasonably introductory but it isn't dumbed-down (thankfully). Especially appreciated the first section of the book on electronic trading basics and economic/portfolio fundamentals. Go deeper and such an amazing repertoire of models (and ideas) this book will keep you occupied for many months. It covers a huge amount and used in conjunction with the internet for reference you'll find you can gain significant momentum to move forward beyond your expectations. I knocked off a star as much of the environment (uses a HUGE number of libraries) simply causes conflicts of library versions, and it takes quite a bit of time to sort out these conflicts (even with Conda, most of the issues surround early Pandas dependency with Zipline that breaks other software). Additionally much of the code simply does not work (data files ceasing to exist on the net without extensive searching though you will usually find after searching)... you may need to hack the code to work in many cases, it does assume in my opinion the following: Basic Python Reasonably proficient Pandas (uses quite a bit in some interesting ways) Ability to read-around a lot of the discussed signals/indicators and other economic stuff (is interesting though!) The ability to get up to speed (without guidance) on third-party libraries Ability to configure multiple Conda environments and sort out library issues (stack overflow etc.) Familiarity with basic Regression Understanding of Gradient Boosting Understanding of basic FFNN, Convolutional, NLP, Time-Series, GAN"s. Some Linear Algebra, Calculus and Stats useful (not essential though but will deepen your understanding). Advice: Separate Conda environment (Python 3.6) for Zipline, and a 3.8 for everything else. Still highly recommended in my opinion, but certainly not for the faint of heart and you will need a little experience in resolving stuff like package conflicts, the amount of valuable information in this book however is HUGE, it really seems in a class of its own in a pragmatic sense. Note: Thankfully zipline (from the defunct Quantopian) is not extensively used, mostly in the early chapters as it is garbage (imho), one of the most poorly designed quant trading packages out there, now abandoned by its former owner. It simply is not good software by a long stretch and I suspect very strongly will be dropped in the next book edition, it is kludge upon kludge upon kludge and its only real use was Quantopian integration. That the new owner Robinhood also seems to not care much about it speaks volumes. This may change but I highly doubt it. However it is useful to know the basics as these can help when learning other libraries.

I received a review copy from the publisher. This book is a great guide for the quant-trading enthusiast. If you are new to trading and ML, and are looking to get your hands dirty, I would warmly recommend this introduction. The book combines an introduction to quantitative trading with an introduction to machine learning. It introduces a range of free datasources and open source tools to help the reader get started with building their own models. On the ML side, it provides a tour of the popular ML algorithms from OLS to GANs. It often demonstrates the use of the introduced techniques by implementing a strategy based on them, which is a really nice idea in my view. The reader will emerge with a good understanding of how to build and test a strategy utilizing ML. As a disclaimer, this is not a comprehensive "How to start a hedgefund" guide, as it leaves out many important aspects of running a fund (it is a bit shallow on risk management, leaves out topics like liquidity & cash management, and does not describe how to set up live trading, deal with compliance, etc.)

The actual quality of the physical book is unfortunately poor and some of my pages are coming apart from the spine, with very little use.

El libro muy bien, pero el plazo de entrega muy mal. Hice la compra porque me aseguraron que llegaba un día y llegó 3 días más tarde a pesar de ser una compra Prime.

C'est le top absolu!

Machine Learning for Algorithmic Trading" by Stefan Jansen is, without a doubt, a 5-star resource. It stands out as a rare masterpiece that successfully bridges the gap between rigorous academic theory and practical, hands-on application. Stefan has created a desk reference that everyone in the quantitative finance space needs to own. What sets this book apart is its holistic approach: * The Triad of Learning: It seamlessly integrates complex mathematical theory, deep machine learning concepts, and production-ready Python code. * Deep Dives: It doesn't just scratch the surface; it offers a profound exploration of both the mechanics of ML models and the nuances of algorithmic trading strategies. A Note on Timing: The only observation worth noting is the timeline. With the first edition published in 2018 and the second in 2020, the content predates the explosive rise of the modern Generative AI and Transformer revolution (GPT-3, LLMs, etc.). While these topics are touched upon, the landscape has shifted so drastically that they now warrant a dedicated volume of their own. The Verdict: Despite the rapid evolution of NLP since 2020, the foundational knowledge regarding time-series analysis, decision trees, and strategy backtesting remains timeless and best-in-class. It leaves one burning question: Stefan, is a 3rd Edition in the works? We would love to see your take on the Transformer revolution in finance. It is incredibly rare to find a resource that balances Theory, Math, and Code so perfectly. Stefan doesn't just explain how to run a model; he explains why it works mathematically and how to apply it to actual market data.

Book is quite over-hyped. Writing style is poor. Unstructured long sentences made the message lost in between. Very hard to follow what authors is trying to say. Some good information about trading domain is given. It is better to follow a trading book & ML book separately.

Good book, I just have to practice now 😁