Summary
Do you want to grasp deep learning technologies quickly and effectively even without any machine learning background?
Do you want to understand many state-of-art deep learning techniques with bare-minimum math?
Do you have obstacles to implement a real-life deep learning projects even with easy to use Keras?
This book will ease these pains and help you learn and grasp deep learning technology from ground zero with many interesting real-world examples implemented in Keras/TensorFlow with simple and intuitive syntax.In this book, you will learn:
* a basic deep learning concepts/theory with bare-minimum math* a deep-dived/well-explained MNIST CNN example so that you can really understand Keras sequential model, how to choose loss, optimizer, metrics in Keras etc.
* how to use a pre-trained model by using transfer learning/fine-tune techniques.
* what are CNN, RNN, Seq2Seq, word embedding, CTC, Auto-encoder, DMN, DQN/DDQN, MCTS, Alphago/Alphazero etc, and how they work.
* How those deep learning technologies are applied to NLP, OCR, Speech, Computer Games etc.
Description
Artificial Intelligence (AI), Machine Learning especially Deep Learning has made tremendous progress in recent years. It starts to spread to all industries.
Quote from Andrew Ng, a famous AI researcher: "AI is the new electricity. About 100 years ago, electricity transformed every major industry. AI has advanced to the point where it has the power to transform every major sector in coming years."
Unless you are a refresh graduated student with AI/deep learning major, many of us do not have a formal machine learning/deep learning training before, so it is time to keep updated with latest technology.
Keras is a very popular, easy to use, yet powerful deep learning framework that promotes a simple and intuitive syntax. But if you do not have much deep learning background, you will find difficulties to really understand the keras codes and have obstacles to implement real-life deep learning projects effectively in keras.This book will help you learn and grasp deep learning technology from ground zero with many interesting real world examples using python/keras/tensorflow. It covers many state-of-art deep learning technologies, e.g.: Convoluational neural network (CNN), Recurrent neural network (RNN), Seq2Seq model, word emedding, Connectionist temporal calssification (CTC ), Auto-encoder, Dynamic Memrory Network (DMN), Deep-Q-learning(DQN/DDQN), Monte Carlo Tree search (MCTS), Alphago/Alphazero etc. The books could also be used as a quick guide on how to use and understand deep learning in the real life.
Readers should have basic knowledge of python, scripting etc. Any constructive feedback is welcome.
Free lifetime upgrade ( for an electronic copy ) as the book has been and will be frequently updated according to readers' feedback. Please feel free to contact the author.
Table of Contents
- Introduction
- What is deep learning
- Deep neural network basic concepts
- Python and NumPy basic
- Deep learning development environments
- MNIST CNN example: A deep dive of how to handle image data
- Pre-trained model, transfer learning and fine-tuning
- Recurrent neural network - how to handle sequences data
- Natural Language Processing
- Optical character recognition
- Audio processing, speech processing
- Autoencoder network
- Deep reinforcement learning
- Learning from scratch (self-play) AlphaZero
- How to deploy deep learning model.
Note
Pytorch version of this book (Pytorch Deep Learning by Example) could be found at: https: //www.amazon.com/dp/B08JKQLB8Z
