Make Your Own Neural Network

I'll second what others have said - blame the store for Kindle typographical errors, not the Author.

Fantastic book, fantastic author. A must read for anyone who deals with computers on a professional level. This technology has far reaching ramifications.

This book should take you maybe a week or two to go through (depending on your familiarity with calculus and matrix operations). It breaks everything down into bite sized chunks so our poor wet ware can absorb it.

Very good book on neural networks. No previous experience needed. I was able to go through the book in ~10 days along with my usual workload. The author explains things very clearly, using very simple terms.

This is perhaps the most illustrative book I've read to date about neural networks. It properly gives a wonderful overview of the major concepts - feed-forward computation, backpropagation, and gradient descent. Really good stuff for any level of reader. A must-have for students!

This is the most effective but gentle introduction to neural networks I've seen. No other authors of NN primers dare to delve into matrices as the basic math for understanding them but Mr. Rashid pulls it off beautifully. There are lots of very helpful diagrams to make the points clear.

Before I start reading the book, I have 100 questions in my mind about ANN. After my second epoch reading the book, I'm having 3 questions left in my mind.

One thing is to have the domain knowledge and another is to know how to transfer it, to other humans being. The author have both, and he clearly state the book goal at the beginning of it, "This book is for anyone who wants to understand what neural network are".

I bought the kindle version, found some erros\typos so I have to chat with customer support so they "push" the update version.

Overall the author doesn't give for granted previous readers knowledge, he explains how to install Anaconda, he include a section about calculus, he walks you step by step about how a Neural Network is made (In supervised learning with python) and he include a example to run a ANN with raspberry pi. So he anticipates multiple reading scenarios.

I found inside the book some knowledge I haven't seen in other guides and all the book code can be found on github, which is great.

1. Anyone who is crying about kindle...sorry. Kindle version working fine on my phone, laptop and pc.

2. Author does an incredible job at explaining what is being done, how, and why.

3. Anyone crying about it not being deep enough or not math heavy enough...go read. It says gentle introduction to NN. Im so sick of these kids who take a course in calculus and think they are now mathemticians. Real skill is needed to explain a math heavy topic without getting math heavy.

4. This book, in a weekend, does a better job at explaining and showing the beginning of NN than udacity, course rate, or edx.

5. I wish this author would put out more books.

I really enjoyed the book.

I was really looking forward to getting the code to run but had a lot of problems with bad URLs, iPython, etc. Pretty much lost interest at that point and moved on to the next think on my do-list. I am sure I could get thing working after a bunch of google searches, FAQ searchers, etc, but iPython left a bad taste so giving up.

At some point I would like to get the code to run.

[UPDATE] After few minor url changes and getting used to iPython, things all fell into place. Updating my rating to 5 !

Content is terrific (as a practical introduction to neural nets) and the price is quite fair, however this "ebook" is a PDF not a kindle or epub book. Meaning it is very irritating to read in an e-reader. The author should take a weekend to make a proper ebook. It is not difficult and this is expected by ebook customers.

Clear. Takes on one application (number recognition) and creates a python application, step by step, illustrating on the way the various elements of neural net theory. one should remember that Python is much slower than other languages, such as Java, or C++, and this approach is good for learning, not for serious applications. The RaspberryPi chapter does not belong in this book, although one can appreciate the playfulness of this exercise.

This is regarding the Kindle edition of the book. It's very inexpensive for a reason. It isn't actually a Kindle book... It's a PDF that you can download onto certain devices that have the Kindle app on them. I can't actually download it to my Kindle... I'll be trying to get my four bucks back. It's unfortunate because I actually do need the book.

While this book certainly covers the very basics of learning what a neural net is, the entire build up of the book is to work with the MNIST dataset. I was disappointed with the lack of serious mathematics in the book, as I had bought the book to understand the more complex theories of neural networks. For more serious mathematicians and computer scientists, there are better resources on the web.

I won't repeat what the other customers already wrote. I can confirm, it is indeed really outstanding.
I am also really impressed how responsive the author is to feedback to his book.
There was one one section in the book that was slightly unclear and could lead to confusion, so I asked him for clarification.
He updated the content immediately and wrote a new blog post. Very impressed-

I was looking for a relatively simple overview of neural networks with examples at a detailed programming level. This book delivers.

There are a couple of mostly unrelated appendixes, one on basic calculus and one on raspberry pi. I even enjoyed flipping through them quickly because of the author's general likeability.

A great read - it makes no assumptions that you have prior machine learning experience, reviews just enough matrix algebra for you to understand the examples and ties the basic concepts of neural networks in a very clear and concise way that anyone can understand. Can't wait to try the python code out.

Great for getting you over the initial learning curve for ANNs - gets your hands dirty with a Python implementation on a real-world problem, gives you the confidence and basic understanding needed to dive into deep learning.

This was a really good introduction even to people with advanced mathematics / engineering background who want to do more advanced study but just want the quick main points. It's easy to see, with an advanced mathematics background how the techniques here are extended to other vector spaces other than just those that live on the cartesian plane

One of the best books on Machine Learning. I really enjoyed the style of the author, truly demystifies complex topics. I have many books on Machine Learning and Neural Networks but I always get lost and hit a wall. I purchased "Python Machine Learning" which starts describing similar topics at the beginning but then the book just moved too fast for me. I didn't know the why and how.

But Tariq's book was amazing, the author really takes his time to explain things mixing text with illustrations to make a point. Usually my attention span with these kind of books is few pages here and there a day, but with this book I couldn't stop. I was on a rhythm , and I felt I could retain the knowledge. When I attempted to read the other books, they all start perfectly on the first chapter, then all of sudden you see code and the authors move too quickly assuming the reader is following along or has some background knowledge. With this book I felt I could indulge in a meaningful dialogue with anyone on the topic, and I could explain things better. After reading few chapters from Tariq’s book, I picked up "Python Machine Learning" and what was amazing is that I was able to follow along and all the gibberish stuff started to make sense. So, I would urge anyone to start reading this book first because it covers all the fundamentals including basic Linear Algebra (Matrices).

I haven’t finished the book yet but I can’t wait to complete it, and then move to more advanced topics/books.

I bought this book for kindle would have rated it higher if it had a table of contents. Also, it only works for the kindle pc app and not the windows 8 app (the one that is touch friendly). I'll rate it again when these issues are fixed. Other than that, the book is good entry point for anyone who wishes to code simple neural networks.

This book does a terrific job of explaining the key concepts that make up a neural network one at a time, building up a program chapter-by-chapter that leads to a practical example of a working machine learning system.

I was pleasantly surprised at how clearly it's all presented and comes together, as a curious novice this opened my eyes to a world of possibilities in artificial intelligence.

Very nice book on understanding the concepts and putting them straight into practice. Was able to read this in a couple of hours and apply the concepts immediately - definitely interested in machine learning now. Very cool!

If you have ever been curious about machine learning, artificial intelligence or expert systems here is a good introduction to artificial neural networks which are a fundamental part of these technologies. The first half of the book describes the theory and in the second half you will learn to code the neural network using the programming language python and jupyter notebooks. I hadn't programmed in over 30 years so it was interesting to see how things had changed since fortran! The book assumes you have no advanced math but it helps to know some elementary linear algebra and differential calculus to understand the theory of neural networks.

Perfect introduction to neural networks and Python. Very beginner (or math rusty) friendly guided tour from beginning to end of the construction of a useful back propagation neural network that can be easily adapted to usefulness beyond the handwritten number recognition system that you finish with.

Some errors in grammar, no big deal. But there are issues with diagrams being labeled incorrectly and referenced continuously. Kind of annoying when the terms in the equation are wrong. The material is explained super easily though so its pretty obvious what the author meant in the cases of these errors.