XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science

XGBoost, Python, Pandas, Feature Engineering, Machine Learning, Data Science, deep learning, NLP,Time Series Forecasting

3.90 (11 reviews)
Udemy
platform
English
language
Data Science
category
instructor
92
students
5 hours
content
Jan 2023
last update
$89.99
regular price

What you will learn

Learn the top skill to become a Machine Learning Engineer or Data Scientist

Learn XGBoost, the best and most popular algorithm for tabular data

Leverage Pandas for Feature Engineering and data Visualization

Understand how to define a machine learning project, going from raw data to a trained model

Learn Gradient Boosting Decision Trees working with realistic datasets and Hands on projects

Learn to apply XGBoost to NLP problems using Deep Learning and TF-IDF features

Project 1: Supervised Regression problem where we predict AirBnB listings prices

Project 2: Binary Classification problem where we work with actual logs of a website visits to predict online conversions

Project 3: Multi Class text Classification. We work with large datasets and more than 200 classes

Project 4: Time series Forecasting with XGBoost

Description

The XGBoost Deep Dive course is a comprehensive program that teaches students the top skills they need to become a Python machine learning engineer or data scientist. The course focuses on using the Python version of XGBoost, the best and most popular algorithm for tabular data, and teaches students how to use it effectively for a variety of machine learning tasks.

Throughout the course, students will learn how to leverage Pandas for feature engineering and data visualization, and will understand how to define a machine learning project, going from raw data to a trained model. They will also learn about gradient boosting decision trees and will work with realistic datasets and hands-on projects to apply their knowledge in a practical setting.

In addition, students will learn how to apply XGBoost to Natural Language Processing (NLP) problems using deep learning (Sentence Transformers) and TF-IDF features.


The course includes five hands-on projects with Python:

  1. A supervised regression problem where students predict Airbnb listing prices.

  2. A binary classification problem where students work with actual logs of website visits to predict online conversions.

  3. A multi-class classification problem where we would predict the credit rating of customers in 3 categories

  4. A multi-class text classification problem where students work with large datasets and more than 200 classes.

  5. A time series forecasting problem where students use XGBoost to make predictions.

By the end of the course, students will have a strong understanding of how to use XGBoost, Pandas and Python and will be able to apply these skills to their own machine learning and data science projects.

Content

Introduction

Why XGBoost?
XGBoost Intuition
Set up work environment
Configure Anaconda Python Environment
Feedback to improve the course!

Supervised Regression - Predict AirBnB Listing Prices with XGBoost

Supervised Regression Project
Section Materials
Dataset Overview
Target Variable Definition
List Categorical and Numeric Features
Numeric Feature Engineering with Pandas
Feature Engineering Categorical Variables
Feature Engineering Date Features
Code Clean-Up
Preprocess data with Pandas
Train/Test Split and Missing Values
One Hot Encoding
XGBoost Parameter Tuning and Model Training

Binary Classification Project

Introduction
Section Materials
Data Preprocessing with Pandas
Feature Engineering with Pandas
Manual Hyperparameter Tuning with XGBoost
Feature Importance and Cross Validation with XGBoost
Model Evaluation with AUC
Precision, Recall and F1-Scoreand Probability Cut-Offs
Choosing ONE Probability Cut-Offs
Automated Hyperparameter Tuning

Multi-Class Classification - Credit Score Classification

Multi Class Classification Project
Section Materials
Multi-Class Classification - Dataset Overview
Feature Engineering with Pandas
Feature Engineering with Pandas - 2
Train XGBoost Model - Credit Score Prediction

Text Multi-Class Classification

Text Multi-Class Classification Project
Section Materials
Data Preprocessing for Text Classification
Feature Engineering and Model Training

Time Series Forecasting with XGBoost

Time Series Forecasting with XGBoost
Section Materials
Time Series Forecasting - Daily Data

Screenshots

XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Screenshot_01XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Screenshot_02XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Screenshot_03XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Screenshot_04

Reviews

Michel
June 4, 2023
Good points: He knows what he is doing. Good tips. Bad points: His accent is a little bit hard to understand some times. The classes are fast, the concepts are introduced very fast in a condenced way what makes it difficult to understand, it could be divided in more sections.
Mukul
January 11, 2023
Very in depth project on Binary Classification, Text and Multi-class classification. knowledgeable instructor and Supervised regression predict project was amazing. Videos were clear, easy to follow and understandable.
Martin
January 9, 2023
The course is very good. I don't think this material, and his approach to the subject, is available anywhere else. Surely, it can help as a prior step before doing some kaggle competitions. I enjoyed the projects a lot but would liked to see a bit more theory presented but a great course nevertheless.

Charts

Price

XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Price chart

Rating

XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Ratings chart

Enrollment distribution

XGBoost Deep Dive w/ Python & Pandas | Hands-on Data Science - Distribution chart
5028430
udemy ID
12/16/2022
course created date
1/26/2023
course indexed date
Bot
course submited by