教學大綱表 (112學年度 第2學期)
請遵守智慧財產權,勿使用非法影印教科書,避免觸法。
課程名稱
Course Title
(中文) 深度學習
(英文) Deep Learning
開課單位
Departments
資訊經營學系
課程代碼
Course No.
N4570A
授課教師
Instructor
黃謝璋
學分數
Credit
3.0 必/選修
core required/optional
選修 開課年級
Level
大四
先修科目或先備能力(Course Pre-requisites):
課程概述與目標(Course Overview and Goals): 深度學習是一種「實現機器學習的技術」,能夠利用如人類大腦功能般的「類神經網路」,處理如視覺、聽覺等感知問題,從學習中更新權重與偏向量進行學習,最後進行分類或預測。學會各種神經網路的類型後,教導學生懂得調校神經網路和轉移學習目標,讓讀者能夠真正建構出屬於自己的神經網路模型。
教科書(Textbook) 1. 陳允傑, "TensorFlow與Keras:Python深度學習應用實務",旗標,2019。
參考教材(Reference) 1. Francois Chollet, “Deep Learning with Python,” Manning Publications, 2017.
圖書館電子書(E-book of the Library) 黃日鉦,"人工智慧與深度學習",碁峰資訊
曾吉弘,"實戰AI資料導向式學習",碁峰資訊
藍子軒,"TensorFlow自然語言處理",碁峰資訊
課程大綱 Syllabus 學生學習目標
Learning Objectives
單元學習活動
Learning Activities
學習成效評量
Evaluation
備註
Notes

No.
單元主題
Unit topic
內容綱要
Content summary
1 Understanding Artificial Intelligence and Machine Learning 1. Introduction to Artificial Intelligence
2. Understanding Machine Learning
3. Types of Machine Learning
1. Introduction to Artificial Intelligence
2. Understanding Machine Learning
3. Types of Machine Learning
 
2 Building TensorFlow and Keras Development Environment 1. Understanding TensorFlow and Keras
2. Constructing a Python Deep Learning Environment
3. Creating and Managing Python Virtual Environments
1. Understanding TensorFlow and Keras
2. Constructing a Python Deep Learning Environment
3. Creating and Managing Python Virtual Environments
 
3 Foundations of Deep Learning 1. Fundamentals of Deep Learning
2. Neural Networks in Deep Learning
3. Data in Deep Learning - Tensors
1. Fundamentals of Deep Learning
2. Neural Networks in Deep Learning
3. Data in Deep Learning - Tensors
 
4 1.Neural Network – Multilayer Perceptron (MLP) 2. Building Your Own Neural Network - Multilayer Perceptron Implementation 3.Example of Multilayer Perceptron 1. Learning Process of Neural Networks - Forward and Backward Propagation
2. Activation Functions and Loss Functions
3. Backpropagation Algorithm and Gradient Descent
4. Constructing Neural Networks for Classification Problems
5. Constructing Neural Networks for Regression Problems
6. Saving and Loading Neural Network Models
7. Multiclass Classification of Iris Flower Dataset
8. Survival Analysis of Titanic Dataset
1. Learning Process of Neural Networks - Forward and Backward Propagation
2. Activation Functions and Loss Functions
3. Backpropagation Algorithm and Gradient Descent
4. Constructing Neural Networks for Classification Problems
5. Constructing Neural Networks for Regression Problems
6. Saving and Loading Neural Network Models
7. Multiclass Classification of Iris Flower Dataset
8. Survival Analysis of Titanic Dataset
 
5 Convolutional Neural Network (CNN) 1. Convolutional Operation and Pooling Operation
2. Convolutional Neural Network
3. Pooling Layer and Dropout Layer
1. Convolutional Operation and Pooling Operation
2. Convolutional Neural Network
3. Pooling Layer and Dropout Layer
 
6 Convolutional Neural Network (CNN) 1. Convolutional Operation and Pooling Operation
2. Convolutional Neural Network
3. Pooling Layer and Dropout Layer
1. Convolutional Operation and Pooling Operation
2. Convolutional Neural Network
3. Pooling Layer and Dropout Layer
 
7 Convolutional Neural Network Practical Examples 1. Recognizing Color Images from the Cifar-10 Dataset
2. Removing Image Noise Using Autoencoders
1. Recognizing Color Images from the Cifar-10 Dataset
2. Removing Image Noise Using Autoencoders
 
8 RNN, LSTM, and GRU Neural Networks 1. Understanding Sequential Data
2. Fundamentals of Natural Language Processing
3. Recurrent Neural Network (RNN)
4. Long Short-Term Memory Neural Network (LSTM)
1. Understanding Sequential Data
2. Fundamentals of Natural Language Processing
3. Recurrent Neural Network (RNN)
4. Long Short-Term Memory Neural Network (LSTM)
 
9 Midterm Report Midterm Report Presenting the Concept of the Project  
10 Building Your Recurrent Neural Network 1. Understanding the Internet Movie Database (IMDb)
2. Data Preprocessing and Embedding Layer
3. Creating IMDb Sentiment Analysis with MLP and CNN
4. Building a Recurrent Neural Network with Keras
5. IMDb Sentiment Analysis using RNN, LSTM, and GRU
6. Combining CNN and LSTM for IMDb Sentiment Analysis
1. Understanding the Internet Movie Database (IMDb)
2. Data Preprocessing and Embedding Layer
3. Creating IMDb Sentiment Analysis with MLP and CNN
4. Building a Recurrent Neural Network with Keras
5. IMDb Sentiment Analysis using RNN, LSTM, and GRU
6. Combining CNN and LSTM for IMDb Sentiment Analysis
 
11 Recurrent Neural Network Implementation Examples 1. Creating Handwritten Digit Recognition with LSTM on MNIST Dataset
2. LSTM Model for Predicting Google Stock Prices
3. News Topic Classification on Reuters Dataset using LSTM
1. Creating Handwritten Digit Recognition with LSTM on MNIST Dataset
2. LSTM Model for Predicting Google Stock Prices
3. News Topic Classification on Reuters Dataset using LSTM
 
12 Data Preprocessing and Data Augmentation 1. Text Data Preprocessing
2. IMDb Network Movie Data Processing
3. Image Loading and Preprocessing
4. Data Augmentation
5. Small Data Image Classification on Cifar-10 Dataset
1. Text Data Preprocessing
2. IMDb Network Movie Data Processing
3. Image Loading and Preprocessing
4. Data Augmentation
5. Small Data Image Classification on Cifar-10 Dataset
 
13 Fine-Tuning Your Deep Learning Model Identifying Overfitting Issues in Models
Avoiding Underfitting and Overfitting
Accelerating Neural Network Training - Choosing Optimizers
Accelerating Neural Network Training - Batch Normalization
Correctly Timing the Model Training Stop
Automatically Saving Best Weights During Model Training
Identifying Overfitting Issues in Models
Avoiding Underfitting and Overfitting
Accelerating Neural Network Training - Choosing Optimizers
Accelerating Neural Network Training - Batch Normalization
Correctly Timing the Model Training Stop
Automatically Saving Best Weights During Model Training
 
14 Pretrained Models and Transfer Learning Keras Built-in Pretrained Models
Using Pretrained Models for Image Classification Prediction
Understanding Transfer Learning
Transfer Learning for MNIST Handwritten Digit Recognition
Transfer Learning with Pretrained Models
Keras Built-in Pretrained Models
Using Pretrained Models for Image Classification Prediction
Understanding Transfer Learning
Transfer Learning for MNIST Handwritten Digit Recognition
Transfer Learning with Pretrained Models
 
15 Functional API and Model Visualization Visualizing Deep Learning Models
Obtaining Neural Layer Information and Intermediate Layer Visualization
Functional API
Shared Layer Models
Multi-Input and Multi-Output Models
Visualizing Deep Learning Models
Obtaining Neural Layer Information and Intermediate Layer Visualization
Functional API
Shared Layer Models
Multi-Input and Multi-Output Models
 
16 Flexible Learning Week - Watch Flipped Education Videos Jetson nano Lab 1
Complete the course requirements on tronclass.
Jetson nano Lab 1
Complete the course requirements on tronclass.
 
17 Flexible Learning Week - Watch Flipped Education Videos Jetson nano Lab 2
Complete the course requirements on tronclass.
Jetson nano Lab 2
Complete the course requirements on tronclass.
 
18 Final Report Final Report Project Presentation Slides and Written Report  
彈性教學週活動規劃

No.
實施期間
Period
實施方式
Content
教學說明
Teaching instructions
彈性教學評量方式
Evaluation
備註
Notes
1 起:2024-06-10 迄:2024-06-23 2.非同步線上課程 Asynchronous online course Completing Flexible Teaching Tasks on tronClass (Asynchronous Teaching, Watching Tutorial Videos, and Uploading Reflections, Reports, or Post-Test) This section constitutes 10% of the semester grade, and the score is assigned based on the achievement rate.


教學要點概述:
1.自編教材 Handout by Instructor:
■ 1-1.簡報 Slids
■ 1-2.影音教材 Videos
□ 1-3.教具 Teaching Aids
■ 1-4.教科書 Textbook
□ 1-5.其他 Other
□ 2.自編評量工具/量表 Educational Assessment
□ 3.教科書作者提供 Textbook

成績考核 Performance Evaluation: 期末考:30%   期中考:30%   彈性教學:10%   平時考:30%  

教學資源(Teaching Resources):
□ 教材電子檔(Soft Copy of the Handout or the Textbook)
■ 課程網站(Website)
課程網站(Website):http://sanmic.ttu.edu.tw/Teach/DL/DL.htm , Tronclass
扣考規定:https://curri.ttu.edu.tw/p/412-1033-1254.php