Book Image

Apache Spark Machine Learning Blueprints

By : Alex Liu
Book Image

Apache Spark Machine Learning Blueprints

By: Alex Liu

Overview of this book

There's a reason why Apache Spark has become one of the most popular tools in Machine Learning – its ability to handle huge datasets at an impressive speed means you can be much more responsive to the data at your disposal. This book shows you Spark at its very best, demonstrating how to connect it with R and unlock maximum value not only from the tool but also from your data. Packed with a range of project "blueprints" that demonstrate some of the most interesting challenges that Spark can help you tackle, you'll find out how to use Spark notebooks and access, clean, and join different datasets before putting your knowledge into practice with some real-world projects, in which you will see how Spark Machine Learning can help you with everything from fraud detection to analyzing customer attrition. You'll also find out how to build a recommendation engine using Spark's parallel computing powers.

Related Content you might be interested in

Current Title:

Small book image
Apache Spark Machine Learning Blueprints
Alex Liu
May, 2016 | 252 pages
No titles found
Table of Contents (18 chapters)
Apache Spark Machine Learning Blueprints
Apache Spark Machine Learning Blueprints
Credits
Credits
About the Author
About the Author
About the Reviewer
About the Reviewer
www.PacktPub.com
www.PacktPub.com
Preface
Preface
Free Chapter
1
Spark for Machine Learning
Spark for Machine Learning
Spark overview and Spark advantages
Spark computing for machine learning
Machine learning algorithms
MLlib
Spark RDD and dataframes
ML workflows and Spark pipelines
ML workflow examples
Spark notebooks
Summary
2
Data Preparation for Spark ML
Data Preparation for Spark ML
Accessing and loading datasets
Data cleaning
Identity matching
Dataset reorganizing
Dataset joining
Feature extraction
Repeatability and automation
Summary
3
A Holistic View on Spark
A Holistic View on Spark
Spark for a holistic view
Methods for a holistic view
Feature preparation
Model estimation
Model evaluation
Results explanation
Deployment
Summary
4
Fraud Detection on Spark
Fraud Detection on Spark
Spark for fraud detection
Methods for fraud detection
Feature preparation
Model estimation
Model evaluation
Results explanation
Deploying fraud detection
Summary
5
Risk Scoring on Spark
Risk Scoring on Spark
Spark for risk scoring
Methods of risk scoring
Data and feature preparation
Model estimation
Model evaluation
Results explanation
Deployment
Summary
6
Churn Prediction on Spark
Churn Prediction on Spark
Spark for churn prediction
Methods for churn prediction
Feature preparation
Model estimation
Model evaluation
Results explanation
Deployment
Summary
7
Recommendations on Spark
Recommendations on Spark
Apache Spark for a recommendation engine
Methods for recommendation
Data treatment with SPSS
Model estimation
Model evaluation
Recommendation deployment
Summary
8
Learning Analytics on Spark
Learning Analytics on Spark
Spark for attrition prediction
Methods of attrition prediction
Feature preparation
Model estimation
Model evaluation
Results explanation
Deployment
Summary
9
City Analytics on Spark
City Analytics on Spark
Spark for service forecasting
Data and feature preparation
Model estimation
Model evaluation
Explanations of the results
Summary
10
Learning Telco Data on Spark
Learning Telco Data on Spark
Spark for using Telco Data
Methods for learning from Telco Data
Data and feature development
Model estimation
Model evaluation
Results explanation
Model deployment
Summary
11
Modeling Open Data on Spark
Modeling Open Data on Spark
Spark for learning from open data
Data and feature preparation
Model estimation
Results explanation
Deployment
Summary
Index
Index

Deploying fraud detection

As discussed before, MLlib supports model exporting to Predictive Model Markup Language (PMML). For the R notebook, it could run on other environments as well as, and with the PMML R package, R models could be exported. Also, it is possible to deploy models for decision making directly on Apache Spark and make results easily available to users. Therefore, we do export some developed models to PMML for this project.

However, in practice, the users of this project will be more interested in rule-based decision making to use some of our insights and also in score-based decision making to prevent frauds.

Here, we will discuss each one of them only briefly as a full deployment for decision making will need an optimization that is not covered in this chapter.

Turning estimated models into rules and scores is not very challenging and could be done under nonSpark platforms. However, Apache Spark makes things easy and fast. The advantage of utilizing Apache Spark is to allow...

Previous Section
End of Section 8
Next Section