XAI: Explainable Artificial Intelligence, Dissecting Artificial Intelligence
 |
Description
Book Introduction
XAI (eXplainable Artificial Intelligence) is a research field that explains the reasons for artificial intelligence's judgment, and its necessity is increasing as artificial intelligence technology expands.
This contrasts with "black box" AI, where even the algorithm's designers cannot explain the rationale behind the AI's decisions. XAI eliminates the uncertainty surrounding AI's decision-making process, thereby increasing its reliability.
This book covers XAI techniques applicable to traditional machine learning techniques as well as those applicable to cutting-edge deep learning models. Because XAI is a technology that infers the reasons behind AI's decision-making, the process of applying the techniques is as important as the theory itself.
Therefore, this book includes example code that was not covered in existing XAI books.
First, you can learn the theory and then follow the code corresponding to the theory to directly check the XAI analysis results without a separate installation process.
This contrasts with "black box" AI, where even the algorithm's designers cannot explain the rationale behind the AI's decisions. XAI eliminates the uncertainty surrounding AI's decision-making process, thereby increasing its reliability.
This book covers XAI techniques applicable to traditional machine learning techniques as well as those applicable to cutting-edge deep learning models. Because XAI is a technology that infers the reasons behind AI's decision-making, the process of applying the techniques is as important as the theory itself.
Therefore, this book includes example code that was not covered in existing XAI books.
First, you can learn the theory and then follow the code corresponding to the theory to directly check the XAI analysis results without a separate installation process.
- You can preview some of the book's contents.
Preview
index
▣ Chapter 1: Opening the Story
1.1.
DARPA's innovation project
1.2. XAI (2016-2021)
1.3. Conditions for doing XAI well
___1.3.1.
Have a thorough understanding of existing machine learning theory
___1.3.2.
Thinking about how to graft the explanation model
1.4.
XAI and Deep Learning XAI using xgboost?
1.5.
Thank you
▣ Chapter 2: Building a Practice Environment
2.1.
Install Python
2.2. Installing PIP
2.3.
Installing TensorFlow
2.4.
Jupyter Notebook
2.4.1.
Verify Tensorflow-GPU installation
▣ Chapter 3: Preparing for XAI Development
3.1.
Understanding Machine Learning
3.2.
A peek inside the black box
3.3.
Understanding the Difference Between Visualization and XAI
▣ Chapter 4: Decision Tree
4.1.
Decision Tree Visualization
4.2.
Finding feature importance
4.3.
Drawing a partial dependence plot (PDP)
4.4. Using XGBoost
___4.4.1. Advantages of XGBoost
___4.4.2. XGBoost is not deep learning
___4.4.3.
Basic principles
___4.4.4.
Parameters
___4.4.5.
Actual Actions and Tips
4.5.
Exercise 1: Pima Indian Diabetes Decision Model
___4.5.1.
Learn
___4.5.2.
Combining explainable models
___4.5.3.
Tuning the model
___4.5.4.
In conclusion
▣ Chapter 5: Proxy Analysis
5.1.
Introduction to Proxy Analysis
___5.1.1.
Global Proxy Analysis
___5.1.2.
Local Surrogate Analysis
5.2. LIME
___5.2.1. Understanding the LIME Algorithm Intuitively
___5.2.2.
Background theory
___5.2.3.
Exercise 2: Applying LIME to Text Data
___5.2.4.
Exercise 3: Applying LIME to Image Data
___5.2.5.
In conclusion
5.3. SHAP (SHapley Additive exPlanations)
___5.3.1.
Background theory
___5.3.2.
Exercise 4: Using Shapley Values in a Sharing Economy Startup
___5.3.3.
Exercise 5: Finding the Determinants of Boston Housing Prices
___5.3.4.
In conclusion
▣ Chapter 6: Filter Visualization
6.1.
Image filter visualization
6.2.
Combining explainable models
___6.2.1.
Convolutional neural networks and filters
6.3.
Building a Convolutional Neural Network
6.4.
Exercise 6: Visualizing Convolutional Neural Networks
___6.4.1.
Visualize input values and compare them with predicted values
___6.4.2.
Filter visualization
6.5.
In conclusion
▣ Chapter 07: LRP (Layer-wise Relevance Propagation)
7.1.
Background theory
___7.1.1.
Decomposition
___7.1.2.
propagation of validity
7.2.
Exercise 7: Unpacking Convolutional Neural Networks
___7.2.1.
Training a Convolutional Neural Network
___7.2.2.
Obtaining a convolutional neural network subgraph
___7.2.3.
Applying LRP to Convolutional Neural Networks
___7.3. Trends in Deep Learning XAI Before and After the Emergence of LRP
7.4.
In conclusion
▣ Chapter 8: Practical Analysis 1: Decision Trees and XAI
8.1.
Creating AI for Credit Loan Analysis
___8.1.1.
Data Description
___8.1.2.
Column Description
___8.1.3.
Loading data
___8.1.4.
Learning data
8.2. Combining XAI
8.3. Understanding the Model with XAI
8.4. Establishing a Basis for Model Improvement with XAI
▣ Chapter 9: Practical Analysis 2: LRP and XAI
9.1.
Building a Sentiment Analysis Model
___9.1.1.
Data Description
___9.1.2.
Column Description
___9.1.3.
Loading data
___9.1.4.
Learning data
9.2. Combining XAI
9.3. Improving the Original AI with XAI
9.4.
Notice
▣ Chapter 10: Closing the Story
10.1.
Finding Dark Matter
10.2.
Adding XAI to existing models
10.3. The Future of XAI
▣ Chapter 11: References
11.1. Installing the XAI Practice Library
___11.1.1.
Install Python
___11.1.2.
Installing Python Libraries
___11.1.3.
Installing TensorFlow
11.2.
Candlestick chart
11.3.
confusion matrix
___11.3.1.
Accuracy
___11.3.2.
Precision
___11.3.3.
Sensitivity (or Recall)
___11.3.4.
Specificity
___11.3.5.
Fallout rate
___11.3.6.
F1-score
11.4.
TensorFlow Slim
11.5.
Normalization
1.1.
DARPA's innovation project
1.2. XAI (2016-2021)
1.3. Conditions for doing XAI well
___1.3.1.
Have a thorough understanding of existing machine learning theory
___1.3.2.
Thinking about how to graft the explanation model
1.4.
XAI and Deep Learning XAI using xgboost?
1.5.
Thank you
▣ Chapter 2: Building a Practice Environment
2.1.
Install Python
2.2. Installing PIP
2.3.
Installing TensorFlow
2.4.
Jupyter Notebook
2.4.1.
Verify Tensorflow-GPU installation
▣ Chapter 3: Preparing for XAI Development
3.1.
Understanding Machine Learning
3.2.
A peek inside the black box
3.3.
Understanding the Difference Between Visualization and XAI
▣ Chapter 4: Decision Tree
4.1.
Decision Tree Visualization
4.2.
Finding feature importance
4.3.
Drawing a partial dependence plot (PDP)
4.4. Using XGBoost
___4.4.1. Advantages of XGBoost
___4.4.2. XGBoost is not deep learning
___4.4.3.
Basic principles
___4.4.4.
Parameters
___4.4.5.
Actual Actions and Tips
4.5.
Exercise 1: Pima Indian Diabetes Decision Model
___4.5.1.
Learn
___4.5.2.
Combining explainable models
___4.5.3.
Tuning the model
___4.5.4.
In conclusion
▣ Chapter 5: Proxy Analysis
5.1.
Introduction to Proxy Analysis
___5.1.1.
Global Proxy Analysis
___5.1.2.
Local Surrogate Analysis
5.2. LIME
___5.2.1. Understanding the LIME Algorithm Intuitively
___5.2.2.
Background theory
___5.2.3.
Exercise 2: Applying LIME to Text Data
___5.2.4.
Exercise 3: Applying LIME to Image Data
___5.2.5.
In conclusion
5.3. SHAP (SHapley Additive exPlanations)
___5.3.1.
Background theory
___5.3.2.
Exercise 4: Using Shapley Values in a Sharing Economy Startup
___5.3.3.
Exercise 5: Finding the Determinants of Boston Housing Prices
___5.3.4.
In conclusion
▣ Chapter 6: Filter Visualization
6.1.
Image filter visualization
6.2.
Combining explainable models
___6.2.1.
Convolutional neural networks and filters
6.3.
Building a Convolutional Neural Network
6.4.
Exercise 6: Visualizing Convolutional Neural Networks
___6.4.1.
Visualize input values and compare them with predicted values
___6.4.2.
Filter visualization
6.5.
In conclusion
▣ Chapter 07: LRP (Layer-wise Relevance Propagation)
7.1.
Background theory
___7.1.1.
Decomposition
___7.1.2.
propagation of validity
7.2.
Exercise 7: Unpacking Convolutional Neural Networks
___7.2.1.
Training a Convolutional Neural Network
___7.2.2.
Obtaining a convolutional neural network subgraph
___7.2.3.
Applying LRP to Convolutional Neural Networks
___7.3. Trends in Deep Learning XAI Before and After the Emergence of LRP
7.4.
In conclusion
▣ Chapter 8: Practical Analysis 1: Decision Trees and XAI
8.1.
Creating AI for Credit Loan Analysis
___8.1.1.
Data Description
___8.1.2.
Column Description
___8.1.3.
Loading data
___8.1.4.
Learning data
8.2. Combining XAI
8.3. Understanding the Model with XAI
8.4. Establishing a Basis for Model Improvement with XAI
▣ Chapter 9: Practical Analysis 2: LRP and XAI
9.1.
Building a Sentiment Analysis Model
___9.1.1.
Data Description
___9.1.2.
Column Description
___9.1.3.
Loading data
___9.1.4.
Learning data
9.2. Combining XAI
9.3. Improving the Original AI with XAI
9.4.
Notice
▣ Chapter 10: Closing the Story
10.1.
Finding Dark Matter
10.2.
Adding XAI to existing models
10.3. The Future of XAI
▣ Chapter 11: References
11.1. Installing the XAI Practice Library
___11.1.1.
Install Python
___11.1.2.
Installing Python Libraries
___11.1.3.
Installing TensorFlow
11.2.
Candlestick chart
11.3.
confusion matrix
___11.3.1.
Accuracy
___11.3.2.
Precision
___11.3.3.
Sensitivity (or Recall)
___11.3.4.
Specificity
___11.3.5.
Fallout rate
___11.3.6.
F1-score
11.4.
TensorFlow Slim
11.5.
Normalization
Detailed image
Publisher's Review
★ What this book covers ★
◎ Feature Importance
◎ Partial dependence plot
◎ Building an XGBoost model
◎ LIME (Local Interpretable Model-agnostic Explanations)
◎ SHAP (SHapley Additive exPlanations)
◎ Filter visualization
◎ Building a convolutional neural network (CNN)
◎ LRP (Layer-wise Relevance Propagation)
◎ Practical Analysis 1: Building and Explaining a Credit Loan Analysis Model
◎ Practical Analysis 2: Building and Explaining a Photo Sentiment Analysis Model
◎ Feature Importance
◎ Partial dependence plot
◎ Building an XGBoost model
◎ LIME (Local Interpretable Model-agnostic Explanations)
◎ SHAP (SHapley Additive exPlanations)
◎ Filter visualization
◎ Building a convolutional neural network (CNN)
◎ LRP (Layer-wise Relevance Propagation)
◎ Practical Analysis 1: Building and Explaining a Credit Loan Analysis Model
◎ Practical Analysis 2: Building and Explaining a Photo Sentiment Analysis Model
GOODS SPECIFICS
- Date of issue: March 27, 2020
- Page count, weight, size: 340 pages | 175*235*18mm
- ISBN13: 9791158392000
- ISBN10: 1158392001
You may also like
카테고리
korean
korean
![GQ KOREA Mark (Monthly): December [2025]](http://librairie.coreenne.fr/cdn/shop/files/8ef265dbbfbf186523ed75ba7319009d.jpg?v=1765340328&width=3840)
