This text highlights the difference between analytics and data science, using predictive analytic techniques to analyze different historical data, including aviation data and concrete data, interpreting the predictive models, and highlighting the steps to deploy the models and the steps ahead. The book combines the conceptual perspective and a hands-on approach to predictive analytics using SAS VIYA, an analytic and data management platform. The authors use SAS VIYA to focus on analytics to solve problems, highlight how analytics is applied in the airline and business environment, and compare several different modeling techniques. They decipher complex algorithms to demonstrate how they can be applied and explained within improving decisions.
Author(s): Leila Halawi, Amal Clarke, Kelly George
Publisher: Springer
Year: 2022
Language: English
Pages: 158
City: Cham
Contents
Chapter 1: Introduction to Analytics and Data Science
1.1 The Data Analytics and Data Science Revolution
1.2 The Difference Between Data Analytics and Data Science
1.3 Data Analyst Versus Data Scientist
1.4 Example: Data Science and Analytics in Aviation
1.5 Analytics Methods
1.6 Classification of Different Applications and Vendors
1.7 Why the Many Different Methods?
1.8 What You Need to Know About SAS Viya
1.9 Road Plan for the Book
References
Chapter 2: Data Types Structure and Data Preparation Process
2.1 Data Types and Measurements
2.1.1 Taxonomy Data
2.1.2 Big Data
2.2 The Need for Data and Data Sources
2.3 Data Sources
2.4 Data Partitioning and Honest Assessment
2.5 The Necessity of Data Preparation and Curation
2.6 Data Preparation Process
2.7 Exploring SAS VIYA Plaftorm and Import Data
2.8 Understanding the SAS VIYA Interface
2.9 Loading Data into SAS VIYA
Reference
Chapter 3: Data Exploration and Data Visualization
3.1 Aviation Case
3.2 Investigate Phases of Data and Data Exploration
3.3 Putting Descriptive Measures Together
3.4 Descriptive Statistics
3.4.1 Measures of Central Tendency
3.4.2 Measures of Variation
3.4.3 Distribution Shapes
3.5 Data Visualization
3.5.1 Data Visualization Considerations
3.5.2 Types of Data Visualization
3.6 Application Using SAS VIYA
3.7 Answers to Aviation Case Questions
References
Chapter 4: Evaluating Predictive Performance
4.1 The Importance of Evaluating Predictive Performance
4.2 Algorithms
4.3 Evaluating Predictive Performance
4.3.1 Model Assessment
4.3.2 Metrics
4.3.2.1 Metrics for Classification Models
4.3.2.2 Metrics for Regression Models: Numerical Values
4.3.3 Selecting Model Fit Statistics by Prediction Type
4.3.3.1 Fit Statistics
References
Websites
Chapter 5: Decision Trees and Ensemble
5.1 Overview of Decision Trees
5.1.1 Classification Trees
5.1.2 Regression Trees
5.2 Terms Used with Decision Trees
5.3 The Math Behind Decision Trees
5.3.1 Expected Value
5.3.2 Measure of Level of Impurity
5.3.3 Attribute Selection Measures
5.3.3.1 Information Gain
Example
How to Calculate Information Gain
5.3.3.2 Gini Index
GINI Index Example
5.3.3.3 Gain Ratio
5.4 Avoiding Overfitting
5.4.1 Pruning
5.4.2 Concrete Compressive Strength Example 1: Regression Tree Model SAS Visual Analytics
5.5 Ensemble Methods Explained
5.6 Ensemble Methods
5.6.1 Bagging or Bootstrap Aggregation
5.6.1.1 Bootstrap Technique
5.6.1.2 Random Forest
5.6.2 Boosting
5.6.2.1 AdaBoost
5.6.2.2 Gradient Boosting
References
Chapter 6: Regression Models
6.1 Functions and Mathematical Implementation
6.1.1 Functions
6.1.1.1 Types of Functions
6.1.1.2 Linear Function
6.1.2 Coordinate Plane
6.1.2.1 Representation of a Linear Function in Graphical Form Example
6.1.3 Derivative of Functions
6.1.4 Matrices
6.1.5 Definition of Logarithmic Function
6.2 Linear Regression
6.2.1 Applications of Linear Regression
6.3 Multiple Linear Regression
6.3.1 Estimation of the Model Parameters
6.3.2 Concrete Compressive Strength Example 1: SAS Visual Analytics
6.3.3 Concrete Compressive Strength Example 1: Model Builder
6.4 Logistic Regression
6.4.1 Estimating the Coefficients
6.4.2 Types of Logistic Regression
6.4.3 Demonstrations of Logistic Regression: Aviation Example
Reference
Websites
Chapter 7: Neural Networks
7.1 What Are Neural Networks?
7.1.1 How Do Neural Networks Learn?
7.2 The Architecture of Neural Networks
7.2.1 Terminology
7.3 The Mathematics Behind Neural Network
7.3.1 The Common Activation Functions
7.3.2 Limit of Functions
7.3.3 Chain Rule
7.3.4 Working of Neural Network
7.3.4.1 Forward Propagation
7.3.5 Backward Propagation
7.4 Vanishing and Exploding Gradient
7.5 Demonstration for Neural Networks
7.5.1 Concrete Compressive Strength Example 1: SAS Visual Analytics
7.5.2 Demonstrations of Neural Network: Aviation Example
References
Chapter 8: Model Deployment
8.1 Model Deployment
8.1.1 Model Assessment
8.1.2 Model Comparison
8.1.3 Monitoring Model Performance and Updating
8.2 Advantages and Disadvantages of Decision Trees, Regression, Neural Networks, Forest, and Ensemble Models
8.3 Application Example of Model Deployment: Concrete Compressive Strengths
8.4 Conclusion
References
Appendix A: Information for Instructors
SAS Profile and Resources
New to SAS?
SAS Books
SAS VIYA Login
References
Appendix B: Sources of Public Data
Appendix C: Data Dictionary for the Aviation
Data Example
Appendix D: Data Dictionary for the Concrete
Data Example
