Search icon CANCEL
Subscription
0
Cart icon
Your Cart (0 item)
Close icon
You have no products in your basket yet
Arrow left icon
Explore Products
Best Sellers
New Releases
Books
Videos
Audiobooks
Learning Hub
Free Learning
Arrow right icon
Arrow up icon
GO TO TOP
Principles of Data Science

You're reading from   Principles of Data Science Mathematical techniques and theory to succeed in data-driven industries

Arrow left icon
Product type Paperback
Published in Dec 2016
Publisher Packt
ISBN-13 9781785887918
Length 388 pages
Edition 1st Edition
Languages
Arrow right icon
Author (1):
Arrow left icon
Sinan Ozdemir Sinan Ozdemir
Author Profile Icon Sinan Ozdemir
Sinan Ozdemir
Arrow right icon
View More author details
Toc

Table of Contents (15) Chapters Close

Preface 1. How to Sound Like a Data Scientist 2. Types of Data FREE CHAPTER 3. The Five Steps of Data Science 4. Basic Mathematics 5. Impossible or Improbable – A Gentle Introduction to Probability 6. Advanced Probability 7. Basic Statistics 8. Advanced Statistics 9. Communicating Data 10. How to Tell If Your Toaster Is Learning – Machine Learning Essentials 11. Predictions Don't Grow on Trees – or Do They? 12. Beyond the Essentials 13. Case Studies Index

Ensembling techniques

Ensemble learning, or ensembling, is the process of combining multiple predictive models to produce a supermodel that is more accurate than any individual model on its own.

  • Regression: We will take the average of the predictions for each model
  • Classification: Take a vote and use the most common prediction, or take the average of the predicted probabilities

Imagine that we are working on a binary classification problem (predicting either 0 or 1).

# ENSEMBLING

import numpy as np

# set a seed for reproducibility
np.random.seed(12345)

# generate 1000 random numbers (between 0 and 1) for each model, representing 1000 observations
mod1 = np.random.rand(1000)
mod2 = np.random.rand(1000)
mod3 = np.random.rand(1000)
mod4 = np.random.rand(1000)
mod5 = np.random.rand(1000)

Now, we simulate five different learning models that each have about a 70% accuracy, as follows:

# each model independently predicts 1 (the "correct response") if random number was at least 0.3
preds1...
lock icon The rest of the chapter is locked
Register for a free Packt account to unlock a world of extra content!
A free Packt account unlocks extra newsletters, articles, discounted offers, and much more. Start advancing your knowledge today.
Unlock this book and the full library FREE for 7 days
Get unlimited access to 7000+ expert-authored eBooks and videos courses covering every tech area you can think of
Renews at $19.99/month. Cancel anytime
Banner background image