Description
Summary
Grokking Deep Learning teaches you to construct deep finding out neural networks from scratch! In his enticing taste, professional deep finding out professional Andrew Trask presentations you the technology beneath the hood, so that you grok for your self each and every element of coaching neural networks.
Purchase of the print ebook features a loose eBook in PDF, Kindle, and ePub codecs from Manning Courses.
About the Technology
Deep finding out, a department of man-made intelligence, teaches computer systems to be informed by means of The usage of neural networks, era impressed by means of the human mind. On-line textual content translation, self-using automobiles, personalised product suggestions, and digital voice assistants are simply a number of the fun brand new improvements conceivable way to deep finding out.
About the Book
Grokking Deep Learning teaches you to construct deep finding out neural networks from scratch! In his enticing taste, professional deep finding out professional Andrew Trask presentations you the technology beneath the hood, so that you grok for your self each and every element of coaching neural networks. The usage of simplest Python and its math-helping library, NumPy, you can teach your personal neural networks to peer and be mindful photographs, translate textual content into other languages, or even write like Shakespeare! When you are performed, you can be totally ready to transport directly to getting to know deep finding out frameworks.
What’s inside
- The technology in the back of deep learning
- Building and coaching your personal neural networks
- Privacy ideas, together with federated learning
- Tips for proceeding your pursuit of deep learning
About the Reader
For readers with highschool-degree math and intermediate programming talents.
About the Author
Andrew Trask is a PhD scholar at Oxford College and a analysis scientist at DeepMind. Up to now, Andrew was once a researcher and analytics product supervisor at Virtual Reasoning, the place he educated the sector’s biggest synthetic neural community and helped information the analytics roadmap for the Synthesys cognitive computing platform.
Table of Contents
- Introducing deep finding out: why you will have to be told it
- Fundamental ideas: how do machines be told?
- Introduction to neural prediction: ahead propagation
- Introduction to neural finding out: gradient descent
- Learning more than one weights at a time: generalizing gradient descent
- Building your first deep neural community: creation to backpropagation
- How to image neural networks: to your head and on paper
- Learning sign and ignoring noise:creation to regularization and batching
- Modeling chances and nonlinearities: activation functions
- Neural finding out approximately edges and corners: intro to convolutional neural networks
- Neural networks that be mindful language: king – guy + lady == ?
- Neural networks that write like Shakespeare: recurrent layers for variable-duration data
- Introducing computerized optimization: permit’s construct a deep finding out framework
- Learning to write down like Shakespeare: lengthy quick-time period memory
- Deep finding out on unseen information: introducing federated learning
- Where to head from right here: a short lived guide
