Book Image

Mastering Machine Learning with Spark 2.x

By : Michal Malohlava, Alex Tellez, Max Pumperla
Book Image

Mastering Machine Learning with Spark 2.x

By: Michal Malohlava, Alex Tellez, Max Pumperla

Overview of this book

The purpose of machine learning is to build systems that learn from data. Being able to understand trends and patterns in complex data is critical to success; it is one of the key strategies to unlock growth in the challenging contemporary marketplace today. With the meteoric rise of machine learning, developers are now keen on finding out how can they make their Spark applications smarter. This book gives you access to transform data into actionable knowledge. The book commences by defining machine learning primitives by the MLlib and H2O libraries. You will learn how to use Binary classification to detect the Higgs Boson particle in the huge amount of data produced by CERN particle collider and classify daily health activities using ensemble Methods for Multi-Class Classification. Next, you will solve a typical regression problem involving flight delay predictions and write sophisticated Spark pipelines. You will analyze Twitter data with help of the doc2vec algorithm and K-means clustering. Finally, you will build different pattern mining models using MLlib, perform complex manipulation of DataFrames using Spark and Spark SQL, and deploy your app in a Spark streaming environment.

Related Content you might be interested in

Current Title:

Small book image
Mastering Machine Learning with Spark 2.x
Michal Malohlava | Alex Tellez | Max Pumperla
Aug, 2017 | 340 pages
Small book image
Machine Learning at Scale with H2O
David Whiting | Gregory Keys
Jul, 2022 | 396 pages
Small book image
Big Data Analytics with Java
RAJAT MEHTA
Jul, 2017 | 418 pages
Small book image
Machine Learning with Spark
Manpreet Singh Ghotra | Rajdeep Dua
Apr, 2017 | 532 pages
Table of Contents (9 chapters)
Preface
Preface
What this book covers
What you need for this book
Who this book is for
Conventions
Reader feedback
Customer support
Free Chapter
1
Introduction to Large-Scale Machine Learning and Spark
Introduction to Large-Scale Machine Learning and Spark
Data science
The sexiest role of the 21st century – data scientist?
Introducing H2O.ai
What's the difference between H2O and Spark's MLlib?
Data munging
Data science - an iterative process
Summary
2
Detecting Dark Matter - The Higgs-Boson Particle
Detecting Dark Matter - The Higgs-Boson Particle
Type I versus type II error
Spark start and data load
Summary
3
Ensemble Methods for Multi-Class Classification
Ensemble Methods for Multi-Class Classification
Data
Modeling goal
Summary
4
Predicting Movie Reviews Using NLP and Spark Streaming
Predicting Movie Reviews Using NLP and Spark Streaming
NLP - a brief primer
The dataset
Feature extraction
Featurization - feature hashing
Let's do some (model) training!
Super learner
Summary
5
Word2vec for Prediction and Clustering
Word2vec for Prediction and Clustering
Motivation of word vectors
Word2vec explained
Doc2vec explained
Applying word2vec and exploring our data with vectors
Creating document vectors
Supervised learning task
Summary
6
Extracting Patterns from Clickstream Data
Extracting Patterns from Clickstream Data
Frequent pattern mining
Pattern mining with Spark MLlib
Deploying a pattern mining application
Summary
7
Graph Analytics with GraphX
Graph Analytics with GraphX
Basic graph theory
GraphX distributed graph processing engine
Graph algorithms and applications
Summary
8
Lending Club Loan Prediction
Lending Club Loan Prediction
Motivation
Preparation of the environment
Data load
Exploration – data analysis
Summary

Summary

This chapter was all about the binary classification problem: true or false and, for our example, the signal indicative of the Higgs-Boson or background noise? We have explored four different algorithms: single decision tree, random forest, gradient boosted machine, and DNN. For this exact problem, DNNs are the current world-beaters as the models can continue to train for longer (that is, increase the number of epochs) and more layers can be added (http://papers.nips.cc/paper/5351-searching-for-higgs-boson-decay-modes-with-deep-learning.pdf)

In addition to exploring four algorithms and how to perform a grid-search against many hyper-parameters, we also looked at some important model metrics to help you better differentiate between models and understand ways to define how good is good. Our goal for this chapter was to expose you to a variety of different algorithms and...

Previous Section
End of Chapter 2
Next Chapter