XGBoost Master Class in Python
3.8 (35 ratings)
Course Ratings are calculated from individual students’ ratings and a variety of other signals, like age of rating and reliability, to ensure that they reflect course quality fairly and accurately.
185 students enrolled

XGBoost Master Class in Python

A Complete Introduction to XGBoost
3.8 (35 ratings)
Course Ratings are calculated from individual students’ ratings and a variety of other signals, like age of rating and reliability, to ensure that they reflect course quality fairly and accurately.
185 students enrolled
Created by Mike West
Last updated 2/2020
English
English [Auto]
Current price: $13.99 Original price: $19.99 Discount: 30% off
5 hours left at this price!
30-Day Money-Back Guarantee
This course includes
  • 1.5 hours on-demand video
  • 16 downloadable resources
  • Full lifetime access
  • Access on mobile and TV
  • Certificate of Completion
Training 5 or more people?

Get your team access to 4,000+ top Udemy courses anytime, anywhere.

Try Udemy for Business
What you'll learn
  • You'll be able to add your rankings on Kaggle to your resume
  • You'll be able to take what you've learned in the course and apply it to the real world
  • You'll understand the machine learning workflow
  • You'll learn why a class of models known as gradient boosters have taken over competitive modeling
  • You'll learn how to tune an XGBoost model
  • The majority of the course is programmtic with real-world code samples
Requirements
  • A basic understanding of programming in Python
  • Familiarity with the machine learning process
Description

"An in depth course on XGBoost with code, examples and caveats. Very valuable." - Thibaut

"This was a very comprehensive course on the benefits and how to configure the gradient booster XGBoost." Kevin K

"Nice and quick course with concise code examples. I would recommend to someone with a bit of ML experience, not for beginners (as he says in the first lecture)." Alex G.

"Good breadth of coverage on this topic. Good examples and documentation. To elaborate on the who-this-is-for section, if you know machine learning but not XGBoost specifically, this is for you." Louis B

"Great code samples to get started on my own problems. Thanks!" Stephen E.

Welcome to XGBoost Master Class in Python.

My name is Mike West and I'm a machine learning engineer in the applied space. I've worked or consulted with over 50 companies and just finished a project with Microsoft. I've published over 50 courses and this is 49 on Udemy. If you're interested in learning what the real-world is really like then you're in good hands.

There's a big difference between real world machine learning and what you read everywhere else. This course is going to focus on the real world.  The real world is often referred to as applied machine learning.

Important - I want you to take my course, however, I want the course to be right for you. This is NOT an entry level course in machine learning. You'll need have a solid background in core Python. If you haven't taken my courses on data wrangling and core Python for machine learning please do that first.

This is a master class but what does that mean? It simply means the course is more advanced. You'll need to have a solid background in Python and I'd suggest a well rounded background on the fundamentals of machine learning.

In applied machine learning almost all your models will be supervised learning models. That simply means you'll build your models against existing data.  That data will be in the shape of an array.  Think of an excel spreadsheet.  Most models will require your data be in that specific shape before they can model it.

In competitive modeling and the real world, a group of algorithms known as gradient boosters has taken the world be storm.  They've won almost every single competition in the structured data category.  In this course I'm going to show you how to use them to score high on the world's most competitive machine learning competition. Kaggle

Here's what you'll learn:

  • Define gradient boosters

  • Cleanse your data for success

  • Build award winning models with XGBoost

  • Use them on your real world models

  • Add the ranking to your resume

Competitive modeling tells employers you understand the basics of the machine learning workflow.  If you can work through the machine learning workflow from end to end your chances of securing a job in this space are greatly improved.

Make no mistake, the barrier to entry in this space is large.  While this is only one step in a long arduous process to becoming a real world machine learning engineer, it's one of the most important things you can do right now to build your skills and your resume.

If you really want to be a part of one the most exciting career paths in the world then take this course now!!!

Who this course is for:
  • If you want to become a machine learning engineer then this course is for you.
  • If you want something beyond the typical lecture style course then this course is for you.
  • If you want to impress perspective employers with your data science acumen this course is for you
Course content
Expand all 29 lectures 01:20:23
+ Introduction
5 lectures 11:43

Welcome to the course. In this lecture let's find out what the course is about.

Preview 01:22

Is this course right for you? I want you to take my course but more importantly, I want the course to be what you are looking for.

Preview 02:17

Let's learn who created XGBoost and define what XGBoost is.

Preview 02:13

Let's summarize what we covered in this section.

Summary
00:59
+ Gradient Boosting
8 lectures 26:08

In this lecture, let's discuss what will be covered in this section.

Section Introduction
01:19

In this lecture let's learn about the weak learner in gradient boosting, the decision tree.

Gradient Boosting and the Decision Tree
04:15

In this lesson let's learn what recursive binary splitting is and why it's important in gradient boosting.

Recursive Binary Splitting
03:24

In this demo, let's use label encoding to change some text to numbers.

Demo: Label Encoding with XGBoost
05:47

An ensemble is a group of models. Most gradient boosters use an ensemble of decision trees. Let's learn about them in this lesson.

Ensembles
02:56

Bagging and Boosting are often discussed in concert. Let's define what they are in this lesson.

Bagging and Boosting
03:56

In this lecture let's learn how to use the gold standard in segmenting our data for modeling.

Demo: Kfold Cross Validation with XGBoost
02:52

Let's wrap up what was covered in this section.

Summary
01:39
+ Going Deeper with XGBoost
16 lectures 42:32

In this lecture, let's discuss what will be covered in this section.

Section Introduction
01:22

In this lecture you'll see the importance of properly handling missing data.

Demo: Handling Missing Data
04:59

Serialize means to save your model to disk. In this lesson you'll learn how to save and recall your saved model.

Demo: Serialize a Model with Pickle
02:28

Let's find out how important your features are.

Demo: Importance Scores using XGBoost
01:14

Warning. Use the XGBoost feature scores at your own risk.

Demo: Caution using Importance Scores in XGBoost
03:57

Let's monitor the performance of an XGBoost model.

Demo: Monitor Model Performance
02:40

Let's use learning curves to gain more insight into our models.

Demo: Model Evaluation uisng Learning Curves
01:59

In this lecture, you'll learn about early stopping and how to use it.

Demo: Early Stopping in XGBoost
01:36

We've built quite a few classification models. Let's build a regression model using XGBoost.

Demo: Regression Model in XGBoost
02:42

Let's check to see if XGBoost is indeed using parallelism.

Demo: Parallelism in XGBoost
01:56

In this lecture, let's find out what SciKit-Learn and XGBoost have to say about the default hyperparameters.

Demo: Hyperparameter Default Recommendations
03:11

In this lecture let's tune the number of trees to find out the correct number for our build.

Demo: Tuning the Number of Decision Trees
02:47

Row subsampling involves selecting a random sample of the training dataset without replacement. Let's see it in action in this lesson.

Demo: Tuning Row Subsampling
01:37

In this lecture let's tweak the learning rate of a model.

Demo: Tuning the Learning Rate
01:57

In this lecture let's create a worthy resume bullet point for the Kaggle Titanic competition.

Demo: Kaggle Top Titanic Model
06:22

Let's wrap up what was learned in this section.

Summary
01:45