Artificial Intelligence with Microsoft Power BI

Spis treści ebooka

• Preface
  • What Is the Current State of AI Technology in Businesses?
  • The Structure of This Book
  • Why Did We Write This Book?
  • Who Is This Book For?
  • Conventions Used in This Book
  • Using Code Examples
  • OReilly Online Learning
  • How to Contact Us
  • Acknowledgments
• 1. Getting Started with AI in the Enterprise: Your Data
  • Overview of Power BI Data Ingestion Methods
  • Workflows in Power BI That Use AI
    • How Are Dataflows Created?
      • Creating a dataflow by importing a dataset
      • Creating a dataflow by importing/exporting a dataflow
      • Creating dataflows by defining new tables
      • Creating dataflows with linked tables
      • Creating dataflows with computed tables
      • Importing a dataflow model
      • Creating dataflows using a CDM folder
    • Things to Note Before Creating Workflows
    • Streaming Dataflows and Automatic Aggregations
  • Getting Your Data Ready First
    • Getting Data Ready for Dataflows
    • Where Should the Data Be Cleaned and Prepared?
      • Option 1: Clean the data and aggregate it in the source system
      • Option 2: Clean data from a source to a secondary store
  • Real-Time Data Ingestion Versus Batch Processing
    • Real-Time Datasets in Power BI
      • Setting up streaming datasets
      • Ingesting data into Power BI: Push method versus streaming method
    • Batch Processing Data Using Power BI
    • Importing Batch Data with Power Query in Dataflows
    • The Dataflow Calculation Engine
    • Dataflow Options
      • Option 1: Fully managed by Power BI
      • Option 2: Bring your own data lake
      • Option 3: External dataflows
      • Power BI dataflows in Power BI Desktop
    • DirectQuery in Power BI
    • Import Versus Direct Query: Practical Recommendations
    • Premium, Pro, and Free Power BI
  • Summary
• 2. A Great Foundation: AI and Data Modeling
  • What Is a Data Model?
    • What Is a Fact Table?
      • What is a dimension table?
      • Star and snowflake schemas
    • Why Is Data Modeling Important?
    • Why Are Data Models Important in Power BI?
    • Why Do We Need a Data Model for AI?
  • Advice for Setting Up a Data Model for AI
    • Analytics Center of Excellence
    • Earning Trust Through Data Transactions
    • Agile Data Warehousing: The BEAM Framework
  • Data Modeling Disciplines to Support AI
  • Data Modeling Versus AI Models
    • Data Modeling in Power BI
      • Fact tables and dimension tables
      • Modeling relationships in the data model
    • What Do Relationships Mean for AI?
      • Optimizing data storage and data retrieval
      • Tips for improving performance in the data model
        • Sort columns before bringing them into a Power BI data model
        • Only bring the data that you need
        • Be mindful of bidirectional and many-to-many relationships
        • Use your own custom date table
        • Reduce data load on each page
        • Reduce the number of visuals on the page
        • Avoid iterator functions (e.g., SUMX, RANKX)
        • Optimize the data source
      • Why dont we just have one huge table?
    • Flat File Structure Versus Dimensional Model Structure in Power BI
  • Summary
• 3. Blueprint for AI in the Enterprise
  • What Is a Data Strategy?
    • Artificial Intelligence in Power BI Data Visualization
      • The Power BI decomposition tree
      • What is root cause analysis?
      • Power BI key influencers visuals
      • Q&A visual
    • Insights Using AI
    • Automated Machine Learning (AutoML) in Power BI
    • Cognitive Services
    • Data Modeling
      • Q&A
      • Quick Insights
  • Real-World Problem Solving with Data
    • Binary Prediction
      • Confusion matrix
      • Accuracy
      • Per-class precision, recall, and F-1
      • Macro-averaged metrics
      • Micro-averaged metrics
      • Evaluation of highly imbalanced datasets
      • Majority class metrics
      • Random-guess metrics
    • Classification
    • Regression
      • R-squared
      • Root mean square deviation
  • Practical Demonstration of Binary Prediction to Predict Income Levels
    • Gather the Data
    • Create a Workspace
    • Create a Dataflow
    • Model Evaluation Reports in Power BI
      • Accuracy report
      • Training report
  • Summary
• 4. Automating Data Exploration and Editing
  • The Transformational Power of Automation
    • Surviving (and Thriving with) Automation
      • How a data analyst can prepare
      • How a data-driven organization can prepare
    • AI Automation in Power BI
  • AI in Power Query
    • Get Data from Web by Example
    • Demo 4-1: Get Data from Web by Example
    • Add Column from Examples
    • Demo 4-2: Add Column from Examples
    • Data Profiling
    • Demo 4-3: Data Profiling
    • Table Generation
    • Demo 4-4: Table Generation
    • Fuzzy Matching
    • Demo 4-5: Fuzzy Matching
  • Intelligent Data Exploration
  • Quick Insights
    • Demo 4-6: Quick Insights
    • Report Creation
    • Demo 4-7: Report Creation
    • Smart Narrative
  • Summary
• 5. Working with Time Series Data
  • More Than Just Timestamps
    • The Components of a Time Series
    • Changes to a Time Series
    • How Trend Lines Work in Power BI
    • Limitations of Trend Lines
    • Demo 5-1: Exploring Taxi Trip Data
  • Forecasting
    • Forecasting for Business
    • How Forecasting Works
    • Limitations of Forecasting
    • Demo 5-2: Forecasting Taxi Trip Data
  • Anomaly Detection
    • Anomaly Detection for Business
    • How Anomaly Detection Works
    • Limitations of Anomaly Detection
    • Demo 5-3: Anomaly Detection with Taxi Trip Data
  • Summary
• 6. Cluster Analysis and Segmentation
  • Cluster Analysis for Business
  • Segmentation Meets Data Science
  • Preprocessing Data for Cluster Analysis
  • How Cluster Analysis Works in Power BI
    • Limitations of Cluster Analysis
    • Demo 6-1: Cluster Analysis with AirBnB Data
  • Summary
• 7. Diving Deeper: Using Azure AI Services
  • Supporting Data-Driven Decisions with a Data Dictionary
  • What Is Azure AI Services?
  • Accessing Azure AI Services in Power BI
    • Creating an Azure AI Services Resource
    • Creating a Power BI Report
  • OpenAI ChatGPT and Power BI
    • What Is the Purpose of the Exercise?
    • Exercise Prerequisites
    • Azure OpenAI and Power BI Example
    • Generating a Secret Key and Code from the OpenAI Website
    • Creating a Streaming Power BI Dataset
    • Dashboard Didnt Work?
  • Summary
• 8. Text Analytics
  • Custom Models Versus Pretrained Models
  • Text as Data
    • Limitations of Text Analytics
    • Demo 8-1: Ingest AirBnB Data
  • Language Detection
    • How It Works
    • Performance and Limitations
    • Demo 8-2: Language Detection
  • Key Phrase Extraction
    • How It Works
    • Performance and Limitations
    • Demo 8-3: Key Phrase Extraction
  • Sentiment Analysis
    • How It Works
    • Recommendations and Limitations
    • Demo 8-4: Sentiment Analysis
    • Demo 8-5: Exploring a Report with Text Analytics
  • Summary
• 9. Image Tagging
  • Images as Data
  • Deep Learning
  • A Simple Neural Network
  • Image Tagging for Business
    • How It Works
    • Limitations of Vision
    • Demo 9-1: Ingest AirBnB Data
    • Demo 9-2: Image Tagging
    • Demo 9-3: Exploring a Report with Vision
  • Summary
• 10. Custom Machine Learning Models
  • AI Business Strategy
    • Organizational Learning with AI
    • Successful Organizational Behaviors
  • Custom Machine Learning
    • Machine Learning Versus Typical Programming
    • Narrow AI Versus General AI
  • Azure Machine Learning
    • Azure Subscription and Free Trial
    • Azure Machine Learning Studio
      • 1. Home
      • 2. Notebooks
      • 3. Automated ML
      • 4. Designer
      • 5. Data
      • 6. Jobs
      • 7. Models
      • 8. Endpoints
      • 9. Compute
      • 10. Data Labeling
    • Demo 10-1: Forecasting Vending Machine Sales
      • Creating an Azure Machine Learning workspace
      • Training a custom model in Azure Machine Learning Studio
      • Deploying a custom model in Azure Machine Learning
      • Consuming a custom model in Power BI
  • Summary
• 11. Data Science Languages: Python and R in Power BI
  • Python Versus R
  • Limitations
  • Setup
    • Setting Up Python
    • Setting Up R
  • Ingestion
    • Ingesting Data with Python
    • Ingesting Data with R
  • Transformation
    • Transforming Data with Python
    • Transforming Data with R
  • Visualization
    • Visualizing Data with Python
    • Visualizing Data with R
  • Machine Learning
    • Using a Pretrained Model with Python on Transform
    • Training a Model with R on Ingest
  • Summary
• 12. Making Your AI Production-Ready with Power BI
  • Strategies to Help Evaluate Models
    • Scenario Without Heteroscedasticity
    • Scenario with Heteroscedasticity
    • How Does Heteroscedasticity Affect AI Models?
    • What Can Be Done If Heteroscedasticity Is Suspected?
  • Making Your AI Model Ready for the Real World
    • Assessing the Costs and Benefits to the Business
    • Example ROI Calculation
  • Can the Business Teams Have Confidence in the AI Model?
    • Is the Model Result Just a Fluke?
    • Assuring Ongoing Model Performance
  • Making Your AI Production-Ready in Power BI
    • Data Lineage for the AI Model
    • Using the Scored Output from the Model in a Power BI Report
  • Summary
• 13. The AI Feedback Loop
  • How Do You Start the Next Project?
  • How Does Feedback Affect the Training and Development of AI Models?
  • AI and Edge Cases in Feedback
  • How Can Feedback Help Fix Errors in an AI Model?
  • AI, Bias, and Fairness
  • Explainable AI and Feedback
  • How Can Members of Organizations Address Ethics and AI?
    • Transfer Learning in Model Training
  • How Are Other Organizations Using the AI Feedback Loop?
  • How Can the AI Feedback Loop Help You?
  • AI and Power BIOver to You!
• Index

pokaż cały spis treści

ukryj recenzje