Make Your Own Neural Network by Tariq Rashid (PDF)

Description

A step-by-step gentle journey through the mathematics of neural networks, and making your own using the Python computer language. Neural networks are a key element of deep learning and artificial intelligence, which today is capable of some truly impressive feats. Yet too few really understand how neural networks actually work.This guide will take you on a fun and unhurried journey, starting from very simple ideas, and gradually building up an understanding of how neural networks work. You won’t need any mathematics beyond secondary school, and an accessible introduction to calculus is also included.The ambition of this guide is to make neural networks as accessible as possible to as many readers as possible – there are enough texts for advanced readers already!You’ll learn to code in Python and make your own neural network, teaching it to recognise human handwritten numbers, and performing as well as professionally developed networks.Part 1 is about ideas. We introduce the mathematical ideas underlying the neural networks, gently with lots of illustrations and examples. Part 2 is practical. We introduce the popular and easy to learn Python programming language, and gradually builds up a neural network which can learn to recognise human handwritten numbers, easily getting it to perform as well as networks made by professionals. Part 3 extends these ideas further. We push the performance of our neural network to an industry leading 98% using only simple ideas and code, test the network on your own handwriting, take a privileged peek inside the mysterious mind of a neural network, and even get it all working on a Raspberry Pi. All the code in this has been tested to work on a Raspberry Pi Zero.

User’s Reviews

Reviews from Amazon users which were colected at the time this book was published on the website:

⭐Who is Tariq Rashid? As I read more and more of this book, finding the answer to this question became paramount. I’ve read a lot of books, both fiction and non-fiction. In fact, I typically read a book a week. But never have I read a text which as so clear, and concise. Such a perfect marriage of simplicity and information about the subject. I actually didn’t want to put the book down! Which is why I had to know about Mr. Rashid. It became clear to me that this man must be among the greatest teachers EVER!What would you normally expect when you buy a book like this? Lots of math? Plenty of dry paragraphs? Definitions on every page? Theorems and proofs? Well, if you do that in this case, you will be disappointed. Instead, what you find is that you will get exactly what you need to understand the topic, and put it into practice. No more, and no less.Mr. Rashid is an absolute master of communicating just what the reader needs to reach the point of understanding. And to get there, he uses a very simple, step-by-step approach. One that builds directly on what has gone before. No hiding the ball. No leaps in logic. Just a simple explanation that leaves no room for doubt (he works through simple examples to hammer needed points home). And at the end of the journey – and indeed, at many points along the way – where are you? I’ll tell you where: you will be saying to yourself “I understand this! I got it! I am so happy I bought this book!”I got an inkling of what was to come when I read the appendices (e.g., about the basics of calculus). I figured the appendices in a book often tell you something about the author – what they decided to include, but not as part of the main text. When I finished them, I wondered why I never had a textbook, or even a teacher, like that? Why couldn’t everyone explain things in this manner? I know when I give a lecture I try to make things interesting and easy to understand, but I suspect I’ve never reached this level, since I’ve never seen anything like it myself. And I’ve had almost 20 years in school!In the end, I have no doubt that I can use what I’ve learned. Indeed, I’ve got some nifty ideas about some ways to try this out, and am looking forward to doing so. So if you’re thinking that the subject might be too difficult for you, I’m here to tell you that if you’ve got high school math, and even if you don’t have a programming background, you can still do this!The bottom line is that if you want to learn about the basics of neural networks, THIS IS THE BOOK TO BUY! Don’t even think about anything else. GET THIS BOOK!Oh yes, so who is Mr. Rashid? He’s a man that is inspired to explain things to others. To really teach them, so that they can work to carry knowledge forward. But he is also a man that is too modest to publish anything about himself in his own books. You can read a little bit about him on amazon.com. That’s what I did. He is truly an inspiration to me. Both in building my confidence in my own ability to learn this subject, and in restoring my confidence in others – proving to me that there are still plenty of people that truly want to do something good in this world.And Mr. Rashid, if you read this, I want you to know that you have not failed. No, you have definitely succeeded, in so many ways. Thank you for taking the time to do such great work with this book. I have never seen anything like it, and I hope you will be inspired to continue in your mission. Technical literature is not my favorite genre, not by a long shot, but in your case, I would be proud to stand first in line to buy whatever you care to write. It’s that good!

⭐The author walks through a Neural network model helping the reader build confidence about this coding process. There is so much more to learn that it is sad when the book concludes.

⭐Are you intrigued at how people learn? Want to learn a bit about how the human brain has neurons that are connected to other neurons and in the process of sending signals between them computation and learning takes place?Do you enjoy a bit of mathematics?Do you appreciate authors who are able to take a complex subject and make it really easy to understand?Have you ever dreamed about writing a program that can learn?Have you ever wanted to write programs that weren’t just hacked together; rather, they were built upon solid mathematical principles?Personally, I answer “yes” to all the above. (Although I am really rusty in mathematics).This book, if read, accomplishes all the above. The book is on neural networks (a.k.a. deep learning). The book is fabulous. I cannot say enough good things about the book and the author. I knew nothing about neural networks prior to reading this book. As I said, my math skills are terribly rusty, but the author works through the math portions with tremendous patience. I am currently on page 92 and haven’t seen any programming stuff. Why? Because the author is first laying down the foundations. Then the programs will be built on top of solid concepts.I highly, highly recommend this book.

⭐I cannot adequately describe my love for this book. I have several years of ML experience but decided recently to re-learn the fundamentals of neural nets for the sake of refreshing that which I have come to overlook or take for granted (this happens, seriously!). What an incredible pleasure it has been to explore this resource! The author does an amazing job of deconstructing the intuition behind neural networks and demystifies the assumptions or fears typically associated with them (probably because he avoids using jargon, which most authors do to veil their lack of understanding of subjects). Do not let past experience with neural nets engender a sense of snobbery when reading a book that might be seen as introductory—I recommend this book for people of all levels of expertise!Any 1 star review is either by those who have purchased the Kindle version (I have not experienced this personally but from the reviews posted would advise people AGAINST purchasing it) or probably by bitter competitors. Contrary to a few acerbic comments I’ve seen, I have no affiliation with the author except for a shared love of both ML and pedagogy. Thank you, Mr. Rashid, for this truly special resource. I will be passing this along to friends and colleague with my highest recommendation!

⭐Within a morning, I had my computer recognising hand-written numerals.If you want to build and run your own simulated neural network and you know the Python programming language, you could probably do the same by downloading the code from Tariq’s Github project webpage. On the other hand if you want a fairly deep understanding of how it all actually works, this book is very good indeed. Machine learning is a subject that seems to be cropping up more and more these days – so if you’re of a vaguely technical mind, why wouldn’t you want to know more about it?Tariq’s enthusiasm for the subject is very catching. It seems the book is aimed at people who understand a bit of maths and are reasonably computer-savvy. You don’t have to be programmer already – but it’d help. He says himself that he wishes that a book like this had been around when he first started looking into the subject – so the book is a boon that saves the average reader from poring over endless, inscrutable academic texts to get into the subject – but at the same time, gives her or him the ability to maybe appreciate those texts afterwards.If you’re new to the concepts or have never written a computer program before then this is likely to be hard work – but still totally doable – and very rewarding.I utterly recommend this book to anyone who is new to and is interested in machine learning or computer programming.Note that it doesn’t work with Kindle devices – you’d need the app or desktop version – or good old fashioned print!

⭐A brilliantly explained introduction to feed forward neural networks which concentrates on the task of image recognition using the well known MNIST set of handwritten digits. The author first explains the theory – walking us through the graphs, matrix theory and calculus needed in “baby steps” which should be understandable to those with shaky math skills. He then teaches enough of the Python programming language to implement this model and the “backprop” algorithm to allow a complete neural network to be implemented. This is done without using any of the special purpose Python libraries such as PyTorch or TensorFlow. Indeed, as a final encore, he shows how even a £5 computer – the. raspberry Pi Zero – can also solve a limited version of the problem. If you supplement the book with Tariq Rashid’s free online blog and GitHub pages together with his YouTube lectures you will have acquired a very good introduction to this new and fascinating area of modern technology. For the price of the kindle book this is an amazing bargain!

⭐Well how neural nets learn using back propagation to be specific. The mathamatics of gradient decent using basic caculus are explained, as are matrix multiplication, input scaling, learning rates, and activation functions. All necessary to really understanding how the networks do what they do.The implementation in Python is amazingly simple making use of a few key thirdparty libaries.A super fun little project, with a few nice twists (such as running the network backwards to see what the network thinks… sort of).

⭐A quote from Carl Satan popped into my mind when I was about halfway through this book : “If you wish to make apple pie from scratch, you must first create the universe”. Rashid’s style is so clear and helpful and the difficulty gradient so well-managed, that you will find yourself looking back over the distance you’ve covered as a mountaineer surveys the lowlands and foothills from the loftiest peak.Some backtracking will probably be required, especially when you get into the equations & calculus, and also the code examples (although these are thoroughly commented) but I was truly impressed by the amount I felt I understood by the time I reached the end.

⭐Fantastic book! I’ve been looking for a book that would take me through the basics of neural networks for sometime. Usually they either gloss over the important stuff so you don’t really get to grips with the basics or they dive straight in and expect you to have a degree in higher maths.This was perfect. The descriptions are clear and well illustrated and the maths, although sometimes a bit tough, is also well explained.The job now is to try and create a network of my own! But at least I feel that I could do it which is something that no other book has given me.

Keywords

Free Download Make Your Own Neural Network in PDF format
Make Your Own Neural Network PDF Free Download
Download Make Your Own Neural Network 2016 PDF Free
Make Your Own Neural Network 2016 PDF Free Download
Download Make Your Own Neural Network PDF
Free Download Ebook Make Your Own Neural Network