Python Machine Learning: A Deep Dive Into Python Machine Learning and Deep Learning, Using Tensor Flow And Keras: From Beginner To Advance

If you have already developed a relatively profound understanding with the python programming language, and are now looking to apply this skill set specifically towards machine and deep learning in finance, this guide is for you. Throughout the book, the author outlines a couple detailed scenarios in which he applies different types of machine and deep learning algorithms towards real life , financial market scenarios to help best bridge the connection between python and how it is being utilized in today's day and age when it comes to artificial intelligence in the financial landscape. I found this book very informative and useful. It is described in the simple language and easy to understand. I got so much new interesting information on all my questions on this topic. I liked this guide and would recommend it for those who need such type of information.

The title is a lie. This is barely an intro into python. It has nothing to do w/ machine or deep learning. It goes from beginner to beginner. Don't waste your time with this book. There are plenty of better books on the market than this.

This book basically consists of 2 parts. 1) how to build and backtest a model driven trading algo and 2) an introduction to python. In that order, which doesn't make sense to me at all. Both sections are done rather poorly. I purchased for section 1 and have found numerous issues with the code provided, some of which will take some time to debug. The content from section 1 seems something that could have come from a blog. Section 2 is basically "why you should learn python", what's a class, module etc, and here's some handy functions you might find useful. The book says you should be comfortable with python before starting, so I'm not sure why this section is included.

This text is really well written, but I would not advise someone who isn't intimately familiar with Python to jump straight into this book. Best python machine learning book for quickly getting up to speed by far!

I keep looking for the rest of the book - there's not enough substance or context. I honestly believed I hadn't downloaded it properly.

Be forewarned, when I downloaded the sample, my Kindle seemed to lock up. Ended up having to reboot my Kindle (hold power button for 5-10 seconds?). I've never had this happen with a Kindle sample/ebook before. Maybe some kind of formatting issue? I notice the formatting is single spaced in the sample, which feels a bit cramped.

Great guide for all who want to learn python machine learning. I got this guide from Amazon. I hope that's the perfect guidebook for all about Python learning. Thanks for the Author.

When I ordered this book, it had more than 60 reviews, 4 and 5 star only. Based on the comments, I was expecting a great introduction to machine learning and the relevant concepts illustrated in python. Instead, I found some reasonable machine learning *examples* in Python with minimal explanation and a lot of filler. It is more like three reasonably long blog articles. While the examples do seem to work, and cover some aspects of machine learning, the explanations are extremely limited and shallow. The author does not go into any depth about why the particular methods are chosen, what they do, or what the drawbacks/tradeoffs/considerations are. The code is poorly formatted for print, nearly every line wraps and is difficult to read. The source code comments are quite good and useful, however. They're actually really essential. The author has one logistic regression example, one backtesting with fnn example, and a longer keras/cnn example. The author has chosen amazon's stock for the examples, which has gone up dramatically over the time period selected (2010-01-01 - 2018-10-29), during which AMZN went up by nearly 10x. The hardcopy is 181 pages long, with only the first 72 pages (minus 3 for the table of contents) focused on machine learning. The 34 page chapter on keras/CNN has more that 20 pages of source code, although this code is very well commented. The remainder of the books is a very shallow introduction to python, for example: Chapter Eleven Functions: What is a function? It is a block of code used to perform a specific task. Through functions, we can break our program. This allows us to have better modularity in favor of organization and better reusability. Then there is a 9 page table listing each of the built-in python functions with a very short description. In summary, while there are reasonably good ML code examples in the book, it is definitely not a thorough explanation of machine learning concepts and nearly 110 out of 181 pages are a very poor python tutorial.

I've learnt a lot with this book, I recommended it without doubt.

Ich habe mich in letzter Zeit etwas mit Machine-Learning mit Python beschäftigt. Nach dem Motto "Kost ja fast nix" habe ich mir auch dieses Buch zugelegt. Ich habe mir nicht viel erwartet und bin diesbezüglich nicht enttäuscht worden. Der Autor entwickelt 2 Modelle. Das erste eine log-lineare Regression mit scikit-learn und das zweite ein CNN-Network mit Tensorflow. Das Logit-Model liefert vom 2017-12-18 bis 2018-10-29 für AMZN einen Gewinn von 28,9%. Das Network von 2017-02-03 weg imposante 70%. Leider verabsäumt der Autor es mit Buy&Hold zu vergleichen. Die AMZN Aktie explodierte von Feb. 2017 von 832.34 bis zum Okt. 2018 auf 1538,88. Buy&Hold hätte seine super-Strategie klar geschlagen. Erstaunlich finde ich, dass der Autor im Okt. 2018 mit dem von ihm verwendeten Python Tool aus dem panda package die Daten von Amazon überhaupt herunter laden konnte. Yahoo hat 2017 zuerst das API geändert und dann überhaupt ausgeschaltet. Das Tool ging im Okt. 2018 nicht mehr (Es gibt inzwischen mit yahoofinancials ein neues Paket das sich auf sehr trickreiche Art und Weise die Daten holt). Kann man vom Kode was Lernen? Der Autor dürfte ein guter Python Programmierer sein. Allerdings ist es schwierig den Kode zu lesen. Das Buch ist A5. Es gehen daher nur sehr wenige Characters in eine Zeile. Außerdem sind Kode und Kommentar im selben Font gesetzt. Wobei die Kommentare vielfach gar nicht direkte Python Comments sind, sondern vom Autor dazugemischte Doku etwa aus dem keras Paket. D.h. es fehlt auch das "#" Zeichen. Er hat sich die Modelle wahrscheinlich nicht selbst ausgedacht, sondern von anderen Quellen übernommen. Wie er selbst anmerkt wurde das CNN-Model für die Bilderkennung entwickelt. Für Zeitreihen verwendet man eher RNN. Ich habe aber - für diesen Zweck - noch kein einfaches RNN Beispiel gefunden. Dieser Punkt hätte mich interessiert. Es ist tröstlich, dass der Autor auch nicht gescheiter ist. Das Buch gliedert sich in 2 Teile. Gut 70 A5-Seiten Modell, der Rest ist eine (unbrauchbare) Python Einführung. Auch wenn das Buch schlecht ist kostete es doch einigen Aufwand es zu verfassen. Angesichts des Preises wird der Autor nicht reich und er präsentiert sich auch nicht als grosser Guru. Den zweiten Stern gab es nach dem Motto "Der Wille gilt fürs Werk".

The book is so and so. Most frustrating part is that there are lines of codes missing at some points. I can not find anywhere authors contact information. Or book code on github.

Once you strip out the "Beginners guide to Python" and the badly laid-out Python code, this book has 21 (yes, TWENTY ONE!) pages about machine learning. Of these, 11 (yes, ELEVEN!!!) of those pages are about Tensorflow and Keras. Quite simply, this is the worst computer book for learning I have ever seen or purchased.