Deep Learning with Swift for Tensorflow: Differentiable Programming with Swift

Bhalley, Rahul

  • 出版商: Apress
  • 出版日期: 2021-02-05
  • 售價: $1,710
  • 貴賓價: 9.5$1,625
  • 語言: 英文
  • 頁數: 287
  • 裝訂: Quality Paper - also called trade paper
  • ISBN: 1484263294
  • ISBN-13: 9781484263297
  • 相關分類: Apple DeveloperDeepLearning 深度學習TensorFlow
  • 下單後立即進貨 (約1週~2週)


Discover more insight about deep learning and how to work with Swift for TensorFlow to develop intelligent apps. TensorFlow was designed for easy adoption by iOS programmers working in Swift. This book covers the established and tested concepts and ties them to modern Swift programming and applicable use in developing for iOS.
Using illustrative examples, the book starts off by introducing you to basic machine learning concepts along with code snippets in Swift for TensorFlow.. Fundamentals of neural networks required to understand today's deep learning research will be covered and put in the context of working in the Swift language with the goal of developing primarily for Apple's mobile ecosystem.

Other important topics covered include computation graphs, loss functions, optimization techniques, regulazrizing nueral networks, recurrent neural networks--such as those used in Siri and Google Translate; and convolutional neural networks. You'll also learn to reuse pre-trained neural networks and work with generative models. Finally, developing and building in security to models is addressed. Swift code will be provided throughout the book to keep the concepts grounded in application within Apple's frameworks.

What You'll Learn

-Write machine learning code in Swift -Run neural networks in Apple environments -Apply fundamental deep learning concepts to mobile app development
Who This Book Is For
Programmers familiar with Swift and the basics of AI


Rahul Bhalley published the first research paper on machine learning in 2016 for an IEEE conference. He actively contributes to open-source works on GitHub, including contributing to others' repositories and writing his own neural networks for generating images. He also focuses on generative models--especially Generative Adversarial Networks and published on the subject in February 2019 with CycleGAN-QP for artist style transfer. He has also worked with Apple's Swift and shares Google's vision of making it easy for others to understand deep learning with Swift.