Low-Code AI Gwendolyn Stripling, Michael Abel

Spis treści książki

• Preface
  • Who Should Read This Book?
  • What Is and Isnt in This Book
  • Conventions Used in This Book
  • Using Code Examples
  • OReilly Online Learning
  • How to Contact Us
  • Acknowledgments
• 1. How Data Drives Decision Making in Machine Learning
  • What Is the Goal or Use Case?
  • An Enterprise ML Workflow
    • Defining the Business Objective or Problem Statement
    • Data Collection
    • Data Preprocessing
    • Data Analysis
    • Data Transformation and Feature Selection
    • Researching the Model Selection or Using AutoML (a No-Code Solution)
    • Model Training, Evaluation, and Tuning
    • Model Testing
    • Model Deployment (Serving)
    • Maintaining Models
  • Summary
• 2. Data Is the First Step
  • Overview of Use Cases and Datasets Used in the Book
    • 1. Retail: Product Pricing
    • 2. Healthcare: Heart Disease Campaign
    • 3. Energy: Utility Campaign
    • 4. Insurance: Advertising Media Channel Sales Prediction
    • 5. Financial: Fraud Detection
    • 6. Energy: Power Production Prediction
    • 7. Telecommunications: Customer Churn Prediction
    • 8. Automotive: Improve Custom Model Performance
  • Data and File Types
    • Quantitative and Qualitative Data
    • Structured, Unstructured, and Semistructured Data
    • Data File Types
    • How Data Is Processed
  • An Overview of GitHub and Googles Colab
    • Use GitHub to Create a Data Repository for Your Projects
      • 1. Sign up for a new GitHub account
      • 2. Set up your projects GitHub repo
    • Using Googles Colaboratory for Low-Code AI Projects
      • 1. Create a Colaboratory Python Jupyter Notebook
      • 2. Import libraries and dataset using Pandas
      • 3. Data validation
      • 4. A little bit of exploratory data analysis
  • Summary
• 3. Machine Learning Libraries and Frameworks
  • No-Code AutoML
  • How AutoML Works
  • Machine Learning as a Service
  • Low-Code ML Frameworks
    • SQL ML Frameworks
      • Googles BigQuery ML
      • Amazon Aurora ML and Redshift ML
    • Open Source ML Libraries
      • AutoKeras
      • Auto-Sklearn
      • Auto-PyTorch
  • Summary
• 4. Use AutoML to Predict Advertising Media Channel Sales
  • The Business Use Case: Media Channel Sales Prediction
    • Project Workflow
    • Project Dataset
  • Exploring the Dataset Using Pandas, Matplotlib, and Seaborn
    • Load Data into a Pandas DataFrame in a Google Colab Notebook
    • Explore the Advertising Dataset
      • Descriptive analysis: Check the data
      • Explore the data
        • Heat maps (correlations)
        • Scatterplots
        • Histogram distribution plot
      • Export the advertising dataset
  • Use AutoML to Train a Linear Regression Model
    • No-Code Using Vertex AI
    • Create a Managed Dataset in Vertex AI
    • Select the Model Objective
    • Build the Training Model
    • Evaluate Model Performance
    • Model Feature Importance (Attribution)
    • Get Predictions from Your Model
  • Summary
• 5. Using AutoML to Detect Fraudulent Transactions
  • The Business Use Case: Fraud Detection for Financial Transactions
    • Project Workflow
    • Project Dataset
  • Exploring the Dataset Using Pandas, Matplotlib, and Seaborn
    • Loading Data into a Pandas DataFrame in a Google Colab Notebook
    • Exploring the Dataset
      • Descriptive analysis
      • Exploratory analysis
    • Exporting the Dataset
  • Classification Models and Metrics
  • Using AutoML to Train a Classification Model
    • Creating a Managed Dataset and Selecting the Model Objective
    • Exploring Dataset Statistics
    • Training the Model
    • Evaluating Model Performance
    • Model Feature Importances
    • Getting Predictions from Your Model
  • Summary
• 6. Using BigQuery ML to Train a Linear Regression Model
  • The Business Use Case: Power Plant Production
  • Cleaning the Dataset Using SQL in BigQuery
    • Loading a Dataset into BigQuery
    • Exploring Data in BigQuery Using SQL
      • Using the Null function to check for null values
      • Using the Min and Max functions to determine acceptable data ranges
      • Saving query results using a DDL statement in BigQuery
  • Linear Regression Models
    • Feature Selection and Correlation
    • Google Colaboratory
    • Plotting Feature Relationships to the Label
    • The CREATE MODEL Statement in BigQuery ML
      • Using the CREATE MODEL statement
      • View evaluation metrics of the trained model
      • Using the ML.PREDICT function to serve predictions
  • Introducing Explainable AI
    • Explainable AI in BigQuery ML
      • Modifying the CREATE MODEL statement
      • Using the ML.GLOBAL_EXPLAIN function
      • Using the ML.EXPLAIN_PREDICT function to compute local explanations
    • Exercises
  • Neural Networks in BigQuery ML
    • Brief Overview of Neural Networks
    • Activation Functions and Nonlinearity
    • Training a Deep Neural Network in BigQuery ML
    • Exercises
  • Deep Dive: Using Cloud Shell to View Your Cloud Storage File
  • Summary
• 7. Training Custom ML Models in Python
  • The Business Use Case: Customer Churn Prediction
  • Choosing Among No-Code, Low-Code, or Custom Code ML Solutions
  • Exploring the Dataset Using Pandas, Matplotlib, and Seaborn
    • Loading Data into a Pandas DataFrame in a Google Colab Notebook
    • Understanding and Cleaning the Customer Churn Dataset
      • Checking and converting data types
      • Exploring summary statistics
      • Exploring combinations of categorical columns
      • Exploring interactions between numeric and categorical columns
    • Transforming Features Using Pandas and Scikit-Learn
      • Feature selection
      • Encoding categorical features using scikit-learn
      • Generalization and data splitting
  • Building a Logistic Regression Model Using Scikit-Learn
    • Logistic Regression
    • Training and Evaluating a Model in Scikit-Learn
    • Classification Evaluation Metrics
    • Serving Predictions with a Trained Model in Scikit-Learn
    • Pipelines in Scikit-Learn: An Introduction
  • Building a Neural Network Using Keras
    • Introduction to Keras
    • Training a Neural Network Classifier Using Keras
  • Building Custom ML Models on Vertex AI
  • Summary
• 8. Improving Custom Model Performance
  • The Business Use Case: Used Car Auction Prices
  • Model Improvement in Scikit-Learn
    • Loading the Notebook with the Preexisting Model
    • Loading the Datasets and the Training-Validation-Test Data Split
    • Exploring the Scikit-Learn Linear Regression Model
    • Feature Engineering and Improving the Preprocessing Pipeline
      • Looking for easy improvements
      • Feature crosses
    • Hyperparameter Tuning
      • Hyperparameter tuning strategies
      • Hyperparameter tuning in scikit-learn
  • Model Improvement in Keras
    • Introduction to Preprocessing Layers in Keras
    • Creating the Dataset and Preprocessing Layers for Your Model
    • Building a Neural Network Model
    • Hyperparameter Tuning in Keras
  • Hyperparameter Tuning in BigQuery ML
    • Loading and Transforming Car Auction Data
    • Training a Linear Regression Model and Using the TRANSFORM Clause
    • Configure a Hyperparameter Tuning Job in BigQuery ML
      • Regularization
      • Using hyperparameter tuning in the CREATE MODEL statement
  • Options for Hyperparameter Tuning Large Models
    • Vertex AI Training and Tuning
    • Automatic Model Tuning with Amazon SageMaker
    • Azure Machine Learning
  • Summary
• 9. Next Steps in Your AI Journey
  • Going Deeper into Data Science
    • Working with Unstructured Data
      • Working with image data
      • Working with text data
    • Generative AI
    • Explainable AI
  • ML Operations
  • Continuous Training and Evaluation
  • Summary
• Index

pokaż cały spis treści

ukryj recenzje