Neural Network Design (2nd Edition)

Seems like an excellent book so farI tried about 10 other books on Neural networks, and always got stuck after the first chapter.This book, actually EXPLAINS stuff.It explains the mathematics very nicely before it is used with Neural networks, and there are loads, and loads of worked examples.As long as you go through the worked examples, you can learn the material.The book refers a lot to Matlab, I dont think you need Matlab to understand the book, but will get more out of the book if you have Matlab.I used Mathematica for symbolics, when studying the book, (I suppose I could have used Matlab instead,but you will need some symbolic software)A very excellent and readable book.They also give it away for free on the internet, but I recommend getting the book, as saves hassle of printing stuff out, and you need a hard copy,to write notes on.A curious thing, an excellent textbook (in my opinion the best on the subject), and apparently they have a set of instructional videos to gowith it. Why don't they just post the instructional videos on youtube, then charge 100 bucks for the book. I would still consider the excellent value at that price. I think the goal of the book, is to prop up Matlab's position as number 1 piece of software for Neural networks.But seriously, a book that helps you learn and study neural networks, and the ONLY book that I have come across so far, that seems to do a good job of explaining.Good explanations, Many, many worked examples, well presented, a LINEAR learning curve.(A lot of books seems to have an easy first chapter to sell the book, then go off the cliff, but this seems linear throughout)So, lots of good points, would recommend, especially if you find, like me you got bogged down in other texts.

This is a fantastic book which introduces various mathematical concepts (vectors, matrices, derivatives, optimization methods), and shows the mathematical derivations of the learning algorithms for several kinds of neural network types (including multi-layer perceptrons and recurrent neural networks).There are several general features of this book which make it a truly superb for going all the way from absolute novice to true expert: (1) Thorough introduction to even the most basic concepts; (2) Numerous illustrations, graphs, and summaries of concepts; (3) Many fully worked-out examples of applying all concepts described in each chapter (very helpful indeed!); and, (4) Advanced examples showing non-trivial and/or realistic applications of methods.If a person with some past exposure to vectors and derivatives wanted to learn about neural networks, from zero to expert, then this would be the very first book I would recommend! In fact, I think a person would do very well if this were their only book until reaching advanced understanding (at which point I would recommend "graduating" to the excellent "Deep Learning" book by Ian Goodfellow, Yoshua Bengio, and Aaron Courville).A minor bittersweet aspect of experiencing this fantastic book is that many other textbooks will seem particularly lame by comparison.Note that this book covers multi-layer perceptrons and recurrent-neural networks in great detail, but does not cover convolutional neural networks (used by most image recognition systems today). However, I think this book represents the best path to being ready to learn about more specialized neural network designs, and I think the extra study required to understand a network like a convolutional neural network would be relatively easy.

This textbook is hands down the most well written Neural Networks textbook available. The notation and illustrations used make this text highly accessible to absolutely anyone interested in the subject - irregardless of your educational background.The content is as self contained as possible, and requires no more than the knowledge of rudimentary calculus as a pre-requisite. The book even includes well written chapters that exhaustively, yet simplistically, cover all the topics from linear algebra and basic mathematical optimization needed to understand the subject of neural networks at its most fundamental level. There is a reason this book gets 4.5 stars on Amazon (but deserves 5!).Here is the primary author's webpage related to the textbook:[...]

This book offers an excellent introduction to the mathematics of neural networks. The authors clearly have designed, built, and tested neural nets for many years. The book is pointed at learners early on in their study of neural nets, as it gives more information about the inner workings of each piece of the NN, and some tested advice about how and when to apply certain techniques. If you'd like to know the details about why NNs work, and how to design your own tools to build NNs, this is the book for you.

I strongly recommend this book for novices, it includes enough details for those without much math background. I personally like the Backpropagation chapter a lot, its the most straightforward explanation I ever read. It is true this book is a little bit out of date(it dose not cover CNN, LSTM or deep learning), however, I still believe these new to this field could benefit from this book.

Very well written en the authors have taken pains to adhere to best pedagogical practices.Money well spent if you:- Have a background in vector/matrix math- Want to understand neural networks deeplySee my comment below, I deducted 1 star for the reason mentioned there.

Worked problems are excellent, but "Exercises" need published answers, too. Book has no index, so looking up specific topics is all but impossible. Discontinuous pagination makes finding a specific item and easily returning to it quite difficult. The review needs to point out that without a copy of MatLab the book's examples are all but intractable. The graphs in the book are utterly illegible.

Questa è la stampa (con rilegatura economica ma accettabile) di un eBook americano disponibile gratuitamente online (con qualche capitolo avanzato in più). Nonostante vi sia un eBook gratis e più completo, a chi affrontasse per la prima volta l'argomento consiglio senz'altro la versione cartacea per non "incrociarsi gli occhi", visto l'alto numero di pagine.Riguardo il contenuto, è uno dei pochissimi testi che abbia mai letto in cui la spiegazione è perfettamente cristallina, sequenziale e interessante. Complimenti agli autori per la chiarezza con cui espongono un argomento per nulla semplice. Sono inoltre fornite tutte le basi matematiche necessarie alla comprensione, per chi non le avesse acquisite nel suo percorso universitario. Testo consigliatissimo, è scritto in modo tale da appassionare all'argomento.Il peso del volume, alto circa 4 cm, dipende in certa parte dai capitoli dedicati alla matematica, oltre che da riassunti, problemi svolti ed esercizi proposti.

This book is a very well written book and by far is the best book in NN design. But the printed version is incomplete and watered down version of the online version. Just search for NNdesign 2nd edition and you'll find the online version which is way more comprehensive compared to this.

Compre este libro porque tenía un precio muy bajo, creo que lo tome con descuento. Me parece un excelente libro para aprender Redes Neuronales y el apoyo usando MathLab es muy bueno. Yo lo uso para mi curso de la Universidad y lo comparto con los Alumnos.

Un muy buen libro sobre Diseño de Redes Neuronales Artificiales.Muy claro al momento de explicar y sus ejemplos son fáciles de seguir.Y, como es costumbre, la entrega fue en tiempo y en forma.

Good book, I will use it in my classes