Predictive Analytics for the Modern Enterprise: A Practitioner's Guide to Designing and Implementing Solutions

✍ Scribed by Nooruddin Abbas Ali

• Publisher: O'Reilly Media
• Year: 2024
• Tongue: English
• Leaves: 619
• Category: Library

✦ Synopsis

The surging predictive analytics market is expected to grow from $10.5 billion today to $28 billion by 2026. With the rise in automation across industries, the increase in data-driven decision-making, and the proliferation of IoT devices, predictive analytics has become an operational necessity in today's forward-thinking companies.

If you're a data professional, you need to be aligned with your company's business activities more than ever before. This practical book provides the background, tools, and best practices necessary to help you design, implement, and operationalize predictive analytics on-premises or in the cloud.

• Explore ways that predictive analytics can provide direct input back to your business
• Understand mathematical tools commonly used in predictive analytics
• Learn the development frameworks used in predictive analytics applications
• Appreciate the role of predictive analytics in the machine learning process
• Examine industry implementations of predictive analytics
• Build, train, and retrain predictive models using Python and TensorFlow

✦ Table of Contents

Preface
Who Is This Book For?
How This Book Is Organized
Conventions Used in This Book
Using Code Examples
O’Reilly Online Learning
How to Contact Us
Acknowledgments
1. Data Analytics in the Modern Enterprise
The Evolution of Data Analytics
Different Types of Data Analytics
Descriptive Analytics
Diagnostic Analytics
Predictive Analytics
Prescriptive Analytics
Knowledge Acquisition, Machine Learning, and the Role of Predictive Analytics
Tools, Frameworks, and Platforms in the Predictive Analytics World
Languages and Libraries
Services
Conclusion
2. Predictive Analytics: An Operational Necessity
The Move from “Data Producing” to “Data Driven”
Challenges to Using Predictive Analytics
People
Data
Technology
Vertical Industry Use Cases for Predictive Analytics
Finance
Healthcare
Automotive
Entertainment
Conclusion
3. The Mathematics and Algorithms Behind Predictive Analytics
Statistics and Linear Algebra
Regression
What Is Regression Analysis?
Regression Techniques
R-squared and P-value
Selecting a Regression Model
Decision Trees
Training Decision Trees
Using Decision Trees to Solve Regression Problems: Regression Trees
Tuning Decision Trees
Other Algorithms
Random Forests
Neural Networks
Support Vector Machines
Naive Bayes Classifier
Other Learning Patterns in Machine Learning
Conclusion
4. Working with Data
Understanding Data
Data Preprocessing and Feature Engineering
Handling Missing Data
Categorical Data Encoding
Data Transformation
Outlier Management
Handling Imbalanced Data
Combining Data
Feature Selection
Splitting Preprocessed Data
Understanding Bias
The Predictive Analytics Pipeline
The Data Stage
The Model Stage
The Serving Stage
Other Components
Selecting the Right Model
Conclusion
5. Python and scikit-learn for Predictive Analytics
Anaconda and Jupyter Notebooks
NumPy in Python
Introduction to NumPy
Generating Arrays
Array Slicing
Array Transformation
Other Array Operations
Exploring a Business Example Using Pandas
Pandas in Python
Import and View Data
Visualize the Data
Data Cleaning and Modification
Reading from Different Data Sources
Data Filtering and Grouping
Scikit-learn
Training and Predicting with a Linear Regression Model
Using a Random Forest Classifier
Training a Decision Tree
A Clustering Example (Unsupervised Learning)
Conclusion
6. TensorFlow and Keras for Predictive Analytics
TensorFlow Fundamentals
Linear Regression Using TensorFlow
Data Preparation
Model Creation and Training
Predictions and Model Evaluation
Deep Neural Networks in TensorFlow
Conclusion
7. Predictive Analytics for Business Problem-Solving
Prediction-Based Optimal Retail Price Recommendations
Using a Simple Linear Regression Model
Using a Polynomial Regression Model
Using Multivariate Regression
An Introduction to Recommender Systems
Building Recommender Systems Using surprise scikit in Python
Credit Card Fraud Classification
Credit Card Fraud Baseline Analysis Using Artificial Neural Networks
Credit Card Fraud Weighted Analysis Using Artificial Neural Networks
Credit Card Analysis with Multiple Hidden Layers in the Artificial Neural Network
Conclusion
8. Exploring AWS Cloud Provider Services for AI/ML
To Cloud or Not to Cloud
Exploring AWS SageMaker
Prerequisites
Data Ingest and Exploration
Data Transformation
Model Training and Prediction
Cleanup
Exploring Amazon Forecast
Import Data
Train the Predictor
Create a Forecast
What-If Analysis
Cleanup
Conclusion
9. Food for Thought
A Few More Use Cases
Navigation and Traffic Management
Credit Scoring
The Social Impact of Predictions
Conclusion
Index
About the Author