Machine Learning in Finance: From Theory to Practice Paperback – 3 July 2020

This book covers a vast number of highly relevant machine learnings topics in an accessible manner (even for non-ML experts) and illustrates their application to finance and other fields with numerous examples in the book and additional exercises or coding applications (Python) for the interested reader.The exposition throughout the books is clear and consistent with plenty of colourful illustrations to reinforce the concepts. The level of prerequisite knowledge is kept to a minimum where possible -- although having an undergraduate degree in maths, physics, statistics, or a related quantitative field will certainly help in studying this book.Being a finance professional with a quantitative background, for me this book provides a deep insight into how ML can be used across various hot topics in quant finance (e.g. algorithmic trading), but also other non-financial disciplines.Impressive work by the authors who showcase their extensive knowledge in the field -- a must buy!

I have a decent understanding of Machine Learning, and wanted to know more about its applications in Finance. It has been a very useful book, as it is rare to find books covering applications of ML in Finance. The best part about this book is that, it also covers various foundational disciplines like Maths & Statistics wherever I felt there was a need for it.I like the fact that it comes with exercises at the end of chapters, and quite a lot of code samples that can be readily executed to understand the concepts. Colour images are a big bonus too.There was a minor issue, 4 or 5 of the colour images, have black text on very dark backgrounds, hence making them unreadable, luckily many of these can be read by executing the code samples, so, it was not a big issue for me. I would have deducted about 0.25 stars for the image issue, but I can't do that, and it is otherwise an excellent addition to my learning.I believe that it will also be equally good for Finance professionals who want to know about Machine Learning, while I belong to the ML group wanting to learn it's applications in Finance.

One slight problem that this book didn't work with my Kindle Paperwhite.

It was a real privilege to be asked to review this book from a delivery and wider team perspective than straight quant finance by my industry peers.As an eighteen year old Physics undergraduate I was taught: you cannot build a house on weak foundations consequently I was taught the mathematical skills I needed and a physicist intuition to be a good Physicist.This textbook will do the same for you in Machine Learning giving you the foundation and intuitions through1. Great academic references2. Clear objectives and conclusions for each chapter3. Many examples for students with complimentary answer book for teachers4. Mathematics presented in an approachable way, which is important to me.Part 1: Machine Learning with Cross sectional dataI like the approach of the book as aim to foundations are laid to provide a solid mathematical framework and intuition to deliver Machine Learning solutions. What Matthew refers to as “mathematical machinery”.Part 2: Sequential LearningLooks at the linkages to finance, econometrics and Machine Learning it aims to address the question “Why do you need to deploy machine learning?” and presents the reasoning behind architectural choices for the deployed – removing the guesswork.Part 3: Sequential Data and Decision MakingLooks at reinforcement learning (RL) and inverse reinforcement learning (IRL) I particularly liked the use case of the “invisible hand” and IRL. You will have to check the book for it.The book finishes with parallels to the Grand Unification theory which for me as a physicist gives me that warm feeling I am in a good space (if you pardon the joke).The book is a great place to build your understanding of the subject with the hype stripped away. From someone who must turn these ideas to business benefits in an organisation I would gladly recommend this book.Satinder

I thoroughly enjoyed reading this book. The authors cover, in great detail, many of the key ML concepts and architectural implementations that are both relevant academically, but also have great practical utility in a financial setting (I speak as a Data Scientist for a large commodity trading outfit). What is particularly helpful, that that the authors solicit and reference a number of successful research papers throughout the book, unlike, for example, many of De Prado's publications whose goal largely appears to be that of proving why everyone else is wrong and simply reference his own work as proof.

The link with the code mentioned in the book is not valid. I tried to contact the site admin she recommended me to find the authors and contact them directly, I also tried to do it via Linkedin, negative outcome. It is an academic source, if I buy the book it has work.Thanks

This very interesting and insightful book presents a very thorough introduction to machine learning in quantitative finance together with a reformulation of some typical quant models and algorithms in the modern context of data driven methods, e.g. establishing connections between Longstaff-Schwartz American Monte Carlo and machine learning. In addition to the standard areas of supervised, unsupervised and reinforcement learning, all capably covered, the book covers more advanced topics like GAILs (generative adversarial imitation learning) and GANs (generative adversarial networks) and provides thorough and up to date bibliography. In addition to finance, the book also touches on topics in microeconomics e.g. structural models for customer behaviour, which has interesting parallels with the section on market microstructure. The Python code resources add practical utility to the theory in this book, which I highly recommend for the serious student, researcher and practitioner in the area.

The content is great but the book arrived damaged. The hard cover of the book is fold and in bad condition.

Now that I've integrated bayesian modeling in my work (as in my presentation "The Right Kind of Volatility" at QuantMinds 2020), I can appreciated more how this book takes its time guiding the reader through the steps of choosing and judging models. The chapters on Reinforcement Learning are more advanced, but worth the time spent learning (I was inspired by the QLBS approach to Black and Scholes). Examples and code make it an outstanding book for those interested in learning more about financial modeling in the 20s.

The authors cut through the hype and rebranding that litters the field of machine learning. They discuss models that are relevant to finance. Most importantly for professionals, they ground their discussion in concepts that will be familiar to statisticians and numerical analysts (quants, in other words).Some of the work presented in this book is new, particularly the sections on inverse reinforcement learning.This will be an excellent resource for a graduate course. Those students who do not have the math background will likely be motivated to get it. There are well-designed Python notebooks that present examples of the analysis. Students with the ability to work with the concepts presented in this book would be welcome in any serious quant shop.

I just started to read the book and I have found it to be very informative for people with interests and background in quantitative finance. Machine Learning, Artificial Intelligence and specially Reinforcement Learning is currently a focus point of research as there has been interesting breakthroughs, e.g. DeepMind's AlphaGo. Financial industry is also benefiting from the machine learning advancements, specially when non-traditional alternative data are available, e.g. sentiment-based trading or natural language processing.The book authors have extensive experience and background in quantitative finance. The book aims to presents the machine learning subject for quantitative finance professionals and graduate students in quantitative disciplines, e.g. Mathematics, Physics, Statistics. The book is divided to three parts: Machine Learning with Cross-Sectional Data, Sequential Learning, and Sequential Data with Decision-Making. Each part encompasses relevant topics presented in a few chapters where each chapters is accompanied by corresponding reference aiding interested readers to dive into the chapter's material. The book is also accompanied with a collection of Python codes to further facilitate the learning process.For readers with knowledge of option pricing, optimal hedging the reinforcement learning part of the book provides the dynamic programming approach toward relevant classical option pricing problems through reinforcement learning closely resembling the celebrated Black-Scholes-Merton model.Overall, the book is valuable resource for Quants to become acquainted with the emerging Machine Learning Applications in Finance. The book should be helpful to the whole Machine Learning, and Artificial Intelligence community, and in particular to quants community in financial industry.

Traditionally finance industry uses mathematical approaches on so-called from "quantitative finance" perspective. Dixon-Halperin-Bilokon's refreshing book does not only capture specialised usage of machine learning in finance but it also serves as a machine learning reference book. They treat chapters in great substance with carefully covering basic concepts in a non-superficial manner.