Home Page Icon
Home Page
Table of Contents for
Cover
Close
Cover
by Eric Boutin, Mike Boyarski, Conor Doherty, Gary Orenstein
Data Warehousing in the Age of Artificial Intelligence
1. The Role of a Modern Data Warehouse in the Age of AI
Actors: Run Business, Collect Data
Applications Producing Data
Enterprise Applications
Operators: Analyze and Refine Operations
Targeting the Appropriate Metric
Accelerating Predictions with ML
The Modern Data Warehouse for an ML Feedback Loop
Dynamic Feedback Loop Between Actors and Operators
2. Framing Data Processing with ML and AI
Foundations of ML and AI for Data Warehousing
AI
ML
Deep Learning
Practical Definitions of ML and Data Science
The Emergence of Professional Data Science
Developing and Deploying Models
Automating Dynamic ML Systems
Supervised ML
Regression
Classification
Unsupervised ML
Cluster Analysis
Online Learning
The Future of AI for Data Processing
The Distributed Era
Advantages of Distributed Datastores
The Future of AI Augmented Datastores
3. The Data Warehouse Has Changed
The Birth of the Data Warehouse
New Performance, Limited Flexibility
The Emergence of the Data Lake
A New Class of Data Warehousing
4. The Path to the Cloud
Cloud Is the New Datacenter
Architectural Considerations for Cloud Computing
Moving to the Cloud
Cost Optimization
Revenue Creation
Choosing the Right Path to the Cloud
5. Historical Data
Business Intelligence on Historical Data
Scalable BI
Query Optimization for Distributed Data Warehouses
Delivering Customer Analytics at Scale
Scale-Out Architecture
Columnstore Query Execution
Intelligent Data Distribution
Examples of Analytics at the Largest Companies
Rise of Data Capture for Analytics
App Store Example
6. Building Real-Time Data Pipelines
Technologies and Architecture to Enable Real-Time Data Pipelines
High-Throughput Messaging Systems
Data Transformation
Operational Datastore
Data Processing Requirements
Memory Optimization
Access to Real-Time and Historical Data
Compiled Query Execution Plans
Multiversion Concurrency Control
Fault Tolerance and ACID Compliance
Benefits from Batch to Real-Time Learning
7. Combining Real Time with Machine Learning
Real-Time ML Scenarios
Supervised and Unsupervised
Continuous and Categorical
Supervised Learning Techniques and Applications
Regression
Categorical: Classification
Determining Whether a Data Point Belongs to a Class by Using Logistic Regression
Unsupervised Learning Applications
Continuous: Real-Time Clustering
Categorical: Real-Time Unsupervised Classification with Neural Networks
8. Building the Ideal Stack for Machine Learning
Example of an ML Data Pipeline
New Data and Historical Data
Model Training
Scoring in Production
Technologies That Power ML
Programming Stack: R, Matlab, Python, and Scala
Analytics Stack: Numpy/Scipy, TensorFlow, Theano, and MLlib
Visualization Tools: Business Intelligence, Graphing Libraries, and Custom Dashboards
Top Considerations
Ingest Performance
Analytics Performance
Distributed Data Processing
9. Strategies for Ubiquitous Deployment
Introduction to the Hybrid Cloud Model
Single Application Stack
Use Case-Centric
Multicloud
On-Premises Flexibility
Hybrid Cloud Deployments
High Availability and Disaster Recovery in the Cloud
Test and Development
Multicloud
Charting an On-Premises-to-Cloud Security Plan
Common Security Requirements
10. Real-Time Machine Learning Use Cases
Overview of Use Cases
Choosing the Correct Data Warehouse
Energy Sector
Goal: Anomaly Detection for the Internet of Things
Approach: Real-Time Sensor Data to Manage Risk
Goal: Take Control of Metering Equipment
Approach: Use Predictive Analytics to Drive Efficiencies
Implementation Outcomes
Thorn
Goal: Use Technology to Help End Child Sexual Exploitation
Approach: ML Image Recognition to Identify Victims
Implementation Outcomes
Tapjoy
Goal: Determine the Best Ads to Serve Based on Previous Behavior and Segmentation
Approach: Real-Time Ad Optimization to Boost Revenue
Implementation Outcomes
Reference Architecture
Datasets and Sample Queries
11. The Future of Data Processing for Artificial Intelligence
Data Warehouses Support More and More ML Primitives
Expressing More of ML Models in SQL, Pushing More Computation to the Database
External ML Libraries/Frameworks Could Push Down Computation
ML in Distributed Systems
Toward Intelligent, Dynamic ML Systems
Search in book...
Toggle Font Controls
Playlists
Add To
Create new playlist
Name your new playlist
Playlist description (optional)
Cancel
Create playlist
Sign In
Email address
Password
Forgot Password?
Create account
Login
or
Continue with Facebook
Continue with Google
Sign Up
Full Name
Email address
Confirm Email Address
Password
Login
Create account
or
Continue with Facebook
Continue with Google
Next
Next Chapter
Data Warehousing in the Age of Artificial Intelligence
Add Highlight
No Comment
..................Content has been hidden....................
You can't read the all page of ebook, please click
here
login for view all page.
Day Mode
Cloud Mode
Night Mode
Reset