Not your usual Succinctly book... &quot;Neural Networks using C# Succinctly&quot; (reg-ware)

Tuesday, June 03, 2014

Not your usual Succinctly book... "Neural Networks using C# Succinctly" (reg-ware)

James D. McCaffrey - Neural Networks using C# Succinctly

I wrote a new book titled “Neural Networks using C# Succinctly”. It was published this week. There are many existing books on neural networks but no good ones (in my opinion) that focus on how to create neural networks from a software developer’s point of view. My book is free and you can download a PDF version from here:

...

I’ve written books before but for “Neural Networks using C# Succinctly” the process was a bit different. I was sitting at my desk one day when I got an unsolicited phone call. Normally I never answer such calls but on this particular day, I did. The call was a young woman named Hilary Bowling who worked for a company called Syncfusion. Hilary asked me if I’d be interested in writing a book about neural networks.

Hilary told me that Syncfusion published relatively short (roughly 100 page) e-books and made them available for free. I was skeptical — I figured there’d have to be a catch of some sort. But in fact, Syncfusion does publish free e-books for software developers. The only minor catch is that you have to register and end up on Syncfusion’s mailing list, but Syncfusion doesn’t take advantage of this (I signed up to see what would happen).

Anyway, it took me a few months to write “Neural Networks using C# Succinctly” (writing a book is much more time-consuming than you might expect) and now it’s available from the Syncfusion Web site...

Syncfusion - Neural Networks using C# Succinctly

Neural networks are an exciting field of software development used to calculate outputs from input data. While the idea seems simple enough, the implications of such networks are staggering—think optical character recognition, speech recognition, and regression analysis. With Neural Networks Using C# Succinctly by James McCaffrey, you'll learn how to create your own neural network to solve classification problems, or problems where the outcomes can only be one of several values. Learn about encoding and normalizing data, activation functions and how to choose the right one, and ultimately how to train a neural network to find weights and bias values that provide accurate predictions.

Table of Contents

Neural Networks

Perceptrons

Feed-Forward

Back-Propagation

Training