Skip to content

sraddhaa/DL-Elegantly

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

499 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

DL-Elegantly

____________________________________________________________________

Table of Contents:

Preface:

This section of the book lays the motivational groundwork for this book, what its intentions are and how to go about comprehending the contents of this book

____________________________________________________________________

Part 1 - Concepts:

This part of the book is a self contained one that explains the history, the concepts, the intuitions and the applications of DL to Machine Vision, Speech, Comprehension, Art, Music and Games

Chapter 1 - Vision

This chapter deals with vision and is focused on conveying a high-level understanding of what deep learning is in the context of vision. The vision analogy provides insight into how deep learning approaches are so powerful and so broadly applicable.

Chapter 2 - Language

This chapter builds atop our deep learning foundations via the vision analogy and examines how deep learning is incorporated into human language applications, with a particular emphasis on how it can automatically learn features that represents the meaning of words.

Chapter 3 - Art

This chapter, introduce some of the concepts that enable deep learning models to create art and intrinsically human, creative activity. It covers the high-level concepts behind GANs, and the novel visual works they can produce. It also establishes the connection between the latent spaces associated with GANs and the word-vector spaces associated with language models.

Chapter 4 - Music

This chapter, introduce some of the concepts that enable deep learning models to create art and intrinsically human, creative activity. It covers the high-level concepts behind GANs, and the novel visual works they can produce. It also establishes the connection between the latent spaces associated with GANs and the word-vector spaces associated with language models.

Chapter 5 - Games

This chapter explains deep reinforcement learning that has produced some of the most surprising artificial-neural-network advances, involving AI breakthroughs of recent years in a very human activity i.e. playing games and interacting with the environment around them. It explains what reinforcement learning (RL) & how RL’s fusion with deep learning has enabled machines to meet or surpass human-level performance on a diverse range of complex challenges in games, and physical-manipulation tasks.

____________________________________________________________________

Part 2: DL - Code & Theory:

Part 1 provided a high-level overview of deep learning we sprinkled with foundational concepts from its hierarchical, representation-learning nature through to its relationship to the field of artificial intelligence. Part II of the book we dive into the low-level theory and mathematics behind in as simple language as possible with iPython notebooks to reify the theory.

Chapter 6 - NN Code

Line-by-line walk-through of a python notebook featuring a neural network model with the theory underpinning the code to make an abstract ideas concrete.

Chapter 7 - NN Theory

Dives into the nitty-gritty theory underlying ANNs introduced in Chapter 6. Explains math basics of artificial neurons, shallow artificial neural networks and deep networks.

Chapter 8 - Shallow NN

We cover, in detail, how individual neural units are linked together to form artificial neural networks and build a feed forward shallow NN to detect images using MNIST data.

Chapter 9 - Deep NN

Introduces gradient descent and back-propagation and applies these concepts s to construct a deep neural network with more than one hidden layer.

Chapter 10 - Advanced Deep NN

Discusses technical issues involved in the creation of high-performing neural networks, techniques that overcome them, applies these in code to architecting a production grade deep neural network.

____________________________________________________________________

Part 3 - DL Applications:

Work through a range of application areas, primarily via hands-on example code including convolutional neural networks and apply them to machine vision tasks, Recurrent neural networks for natural language processing, Generative adversarial networks for visual creativity, Deep reinforcement learning for sequential decision making within complex, changing environments

Chapter 11 - CNN and machine vision application

CNN, Pooling, LeNet5, AlexNet, VGGNet, Tensorflow, Keras, DL Vision App

Chapter 12 - RNN and NLP

Word2Vec, Classification, RNN, LSTM, etc, Keras, TF, NLP application

Chapter 13 - GAN

GAN basics, Dataset, Discriminator, Generator, Adversary, GAN application

Chapter 14: Deep Reinforcement Learning (DRL)

DRL Theory, Q Learning, DQN, OpenAI Gym, SLM Lab Agents

____________________________________________________________________

Part 4 - Architecting Deep Learning Systems:

Chapter 15:

This chapter builds on all of the previous work and presents an end to end System combining Data Science, Architecture, Microservices, Containers, AIOps, AIaaS, DL Frameworks and Innovation

____________________________________________________________________

About

A DL book in Thamizh

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors