Beginning Deep Learning with TensorFlow: Work with Keras, MNIST Data Sets, and Advanced Neural Networks

Long, Liangqu, Zeng, Xiangming

  • 出版商: Apress
  • 出版日期: 2022-01-28
  • 售價: $1,750
  • 貴賓價: 9.5$1,663
  • 語言: 英文
  • 頁數: 740
  • 裝訂: Quality Paper - also called trade paper
  • ISBN: 148427914X
  • ISBN-13: 9781484279144
  • 相關分類: DeepLearningTensorFlow
  • 立即出貨 (庫存 < 3)

商品描述

Incorporate deep learning into your development projects through hands-on coding and the latest versions of deep learning software, such as TensorFlow 2 and Keras. The materials used in this book are based on years of successful online education experience and feedback from thousands of online learners. 

You’ll start with an introduction to AI, where you’ll learn the history of neural networks and what sets deep learning apart from other varieties of machine learning. Discovery the variety of deep learning frameworks and set-up a deep learning development environment. Next, you’ll jump into simple classification programs for hand-writing analysis. Once you’ve tackled the basics of deep learning, you move on to TensorFlow 2 specifically. Find out what exactly a Tensor is and how to work with MNIST datasets. Finally, you’ll get into the heavy lifting of programming neural networks  and working with a wide variety of neural network types such as GANs and RNNs.  

Deep Learning is a new area of Machine Learning research widely used in popular applications, such as voice assistant and self-driving cars. Work through the hands-on material in this book and become a TensorFlow programmer!      

What You'll Learn

• Develop using deep learning algorithms
• Build deep learning models using TensorFlow 2
• Create classification systems and other, practical deep learning applications


Who This Book Is For
Students, programmers, and researchers with no experience in deep learning who want to build up their basic skillsets. Experienced machine learning programmers and engineers might also find value in updating their skills.

商品描述(中文翻譯)

透過實際編碼和最新版本的深度學習軟體(如TensorFlow 2和Keras),將深度學習融入您的開發項目中。本書所使用的材料基於多年的成功線上教育經驗和數千名線上學習者的反饋。

您將從人工智慧的介紹開始,了解神經網絡的歷史以及深度學習與其他機器學習方法的區別。探索各種深度學習框架並建立深度學習開發環境。接下來,您將進入手寫分析的簡單分類程式。一旦掌握了深度學習的基礎知識,您將專注於TensorFlow 2。了解Tensor是什麼,以及如何處理MNIST數據集。最後,您將進入編程神經網絡和使用各種神經網絡類型(如GAN和RNN)的重要部分。

深度學習是機器學習研究的新領域,在流行的應用中廣泛使用,例如語音助手和自駕車。通過本書中的實際材料,成為一名TensorFlow程式設計師!

您將學到什麼:
- 使用深度學習算法進行開發
- 使用TensorFlow 2建立深度學習模型
- 創建分類系統和其他實用的深度學習應用

本書適合對深度學習沒有經驗的學生、程式設計師和研究人員,他們希望建立基本技能。有經驗的機器學習程式設計師和工程師也可能會發現更新他們的技能的價值所在。

作者簡介

​Liangqu Long is a well-known deep learning educator and engineer in China. He is a successfully published author in the topic area with years of experience in teaching machine learning concepts. His two online video tutorial courses “Deep Learning with PyTorch” and “Deep Learning with TensorFlow 2” have received massive positive comments and allowed him to refine his deep learning teaching methods.    

Xiangming Zeng is an experienced data scientist and machine learning practitioner. He has over ten years of experience using machine learning and deep learning models to solve real world problems in both academia and professionally. Xiangming is familiar with deep learning fundamentals and mainstream machine learning libraries such as Tensorflow and scikit-learn.  

作者簡介(中文翻譯)

梁曲龍是中國知名的深度學習教育家和工程師。他在這個領域是一位成功出版的作者,並且擁有多年教授機器學習概念的經驗。他的兩個線上視頻教程《使用PyTorch進行深度學習》和《使用TensorFlow 2進行深度學習》獲得了大量正面評論,並使他能夠改進他的深度學習教學方法。

曾向明是一位經驗豐富的數據科學家和機器學習實踐者。他在學術界和職業生涯中擁有超過十年的使用機器學習和深度學習模型解決實際問題的經驗。向明熟悉深度學習的基礎知識和主流機器學習庫,如Tensorflow和scikit-learn。

目錄大綱

Part 1 Introduction to AI

1. Introduction

1. Artificial Intelligence

2. History of Neural Networks

3. Characteristics of Deep Learning

4. Applications of Deep Learning

5. Deep Learning Frameworks

6. Installation of Development Environment

2. Regression

2.1 Neuron Model

2.2 Optimization Methods

2.3 Hands-on Linear Models

2.4 Linear Regression

3. Classification

3.1 Hand-writing Digital Picture Dataset

3.2 Build a Classification Model

3.3 Compute the Error

3.4 Is the Problem Solved?

3.5 Nonlinear Model

3.6 Model Representation Ability

3.7 Optimization Method

3.8 Hands-on Hand-written Recognition

3.9 Summary

Part 2 Tensorflow

4. Tensorflow 2 Basics

4.1 Datatype

4.2 Numerical Precision

4.3 What is a Tensor?

4.4 Create a Tensor

4.5 Applications of Tensors

4.6 Indexing and Slicing

4.7 Dimension Change

4.8 Broadcasting

4.9 Mathematical Operations

4.10 Hands-on Forward Propagation Algorithm

5. Tensorflow 2 Pro

5.1 Aggregation and Seperation

5.2 Data Statistics

5.3 Tensor Comparison

5.4 Fill and Copy

5.5 Data Clipping

5.6 High-level Operations

5.7 Load Classic Datasets

5.8 Hands-on MNIST Dataset Practice

Part 3 Neural Networks

6. Neural Network Introduction

6.1 Perception Model

6.2 Fully-Connected Layers

6.3 Neural Networks

6.4 Activation Functions

6.5 Output Layer

6.6 Error Calculation

6.7 Neural Network Categories

6.8 Hands-on Gas Consuming Prediction

7. Backpropagation Algorithm

7.1 Derivative and Gradient

7.2 Common Properties of Derivatives

7.3 Derivatives of Activation Functions

7.4 Gradient of Loss Function

7.5 Gradient of Fully-Connected Layers

7.6 Chain Rule

7.7 Back Propagation Algorithm

7.8 Hands-on Himmelblau Function Optimization

7.9 Hands-on Back Propagation Algorithm

8. Keras Basics

8.1 Basic Functionality

8.2 Model Configuration, Training and Testing

8.3 Save and Load Models

8.4 Customized Class

8.5 Model Zoo

8.6 Metrics

8.7 Visualization

9. Overfitting

9.1 Model Capability

9.2 Overfitting and Underfitting

9.3 Split the Dataset

9.4 Model Design

9.5 Regularization

9.6 Dropout

9.7 Data Enhancement

9.8 Hands-on Overfitting

Part 4 Deep Learning Applications

10. Convolutional Neural Network

10.1 Problem of Fully-Connected Layers

10.2 Convolutional Neural Network

10.3 Convolutional Layer

10.4 Hands-on LeNet-5

10.5 Representation Learning

10.6 Gradient Propagation

10.7 Pooling Layer

10.8 BatchNorm Layer

10.9 Classical Convolutional Neural Network

10.10 Hands-on CIFRA10 and VGG13

10.11 Variations of Convolutional Neural Network

10.12 Deep Residual Network

10.13 DenseNet

10.14 Hands-on CIFAR10 and ResNet1

目錄大綱(中文翻譯)

第1部分 人工智慧介紹

1. 介紹
2. 人工智慧
3. 神經網路的歷史
4. 深度學習的特性
5. 深度學習的應用
6. 深度學習框架
7. 開發環境的安裝

第2部分 迴歸

1. 神經元模型
2. 優化方法
3. 線性模型實作
4. 線性回歸

第3部分 分類

1. 手寫數字圖片資料集
2. 建立分類模型
3. 計算誤差
4. 問題是否解決?
5. 非線性模型
6. 模型的表示能力
7. 優化方法
8. 手寫辨識實作
9. 總結

第4部分 Tensorflow

1. Tensorflow 2 基礎
2. 資料型態
3. 數值精度
4. 什麼是 Tensor?
5. 創建 Tensor
6. Tensor 的應用
7. 索引和切片
8. 維度變換
9. 廣播
10. 數學運算
11. 前向傳播演算法實作

第5部分 Tensorflow 2 進階

1. 聚合和分離
2. 資料統計
3. Tensor 比較
4. 填充和複製
5. 資料剪裁
6. 高階操作
7. 載入經典資料集
8. MNIST 資料集實作練習

第6部分 神經網路

1. 神經網路介紹
2. 感知器模型
3. 全連接層
4. 神經網路
5. 激活函數
6. 輸出層
7. 誤差計算
8. 神經網路類別
9. 燃料消耗預測實作

第7部分 反向傳播演算法

1. 導數和梯度
2. 導數的共同特性
3. 激活函數的導數
4. 損失函數的梯度
5. 全連接層的梯度
6. 鏈式法則
7. 反向傳播演算法
8. Himmelblau 函數優化實作
9. 反向傳播演算法實作

第8部分 Keras 基礎

1. 基本功能
2. 模型配置、訓練和測試
3. 儲存和載入模型
4. 自定義類別
5. 模型庫
6. 評估指標
7. 視覺化

第9部分 過擬合

1. 模型能力
2. 過擬合和欠擬合
3. 分割資料集
4. 模型設計
5. 正則化
6. Dropout
7. 資料增強
8. 過擬合實作

第10部分 深度學習應用

1. 卷積神經網路
2. 全連接層的問題
3. 卷積神經網路
4. 卷積層
5. LeNet-5 實作
6. 表示學習
7. 梯度傳播
8. 池化層
9. BatchNorm 層
10. 經典卷積神經網路
11. CIFAR10 和 VGG13 實作
12. 卷積神經網路的變體
13. 深度殘差網路
14. DenseNet
15. CIFAR10 和 ResNet1 實作