This chapter will introduce you to the fundamental idea behind XGBoost—boosted learners. Once you understand how XGBoost works, you'll apply it to solve a common classification problem found in industry: predicting whether a customer will stop being a customer at some point in the future.

Welcome to the course!

Which of these is a classification problem?

Which of these is a binary classification problem?

Introducing XGBoost

XGBoost: Fit/Predict

What is a decision tree?

Decision trees

What is Boosting?

Measuring accuracy

Measuring AUC

When should I use XGBoost?

Using XGBoost

Classification with XGBoost

After a brief review of supervised regression, you'll apply XGBoost to the regression task of predicting house prices in Ames, Iowa. You'll learn about the two kinds of base learners that XGboost can use as its weak learners, and review how to evaluate the quality of your regression models.

Regression review

Which of these is a regression problem?

Objective (loss) functions and base learners

Decision trees as base learners

Linear base learners

Evaluating model quality

Regularization and base learners in XGBoost

Using regularization in XGBoost

Visualizing individual XGBoost trees

Visualizing feature importances: What features are most important in my dataset

Regression with XGBoost

This chapter will teach you how to make your XGBoost models as performant as possible. You'll learn about the variety of parameters that can be adjusted to alter the behavior of XGBoost and how to tune them efficiently so that you can supercharge the performance of your models.

Perché ottimizzare il tuo modello?

Quando è una cattiva idea mettere a punto il tuo modello?

Ottimizzare il numero di boosting rounds

Selezione automatica dei boosting round con early_stopping

Panoramica degli iperparametri di XGBoost

Ottimizzare eta

Ottimizzare max_depth

Ottimizzare colsample_bytree

Ripasso di grid search e random search

Grid search con XGBoost

Ricerca casuale con XGBoost

Limiti della grid search e della random search

Quando dovresti usare grid search e random search?

Fine-tuning your XGBoost model

Take your XGBoost skills to the next level by incorporating your models into two end-to-end machine learning pipelines. You'll learn how to tune the most important XGBoost hyperparameters efficiently within a pipeline, and get an introduction to some more advanced preprocessing techniques.

Review of pipelines using sklearn

Exploratory data analysis

Encoding categorical columns I: LabelEncoder

Encoding categorical columns II: OneHotEncoder

Encoding categorical columns III: DictVectorizer

Preprocessing within a pipeline

Incorporating XGBoost into pipelines

Cross-validating your XGBoost model

Kidney disease case study I: Categorical Imputer

Kidney disease case study II: Feature Union

Kidney disease case study III: Full pipeline

Tuning XGBoost hyperparameters

Bringing it all together

Final Thoughts

Using XGBoost in pipelines

Ames housing prices (preprocessed)

Ames housing prices (original)

Chronic kidney disease

Conosci le basi del supervised learning e vuoi usare modelli all’avanguardia su insiemi di dati reali? Il gradient boosting è oggi una delle tecniche più popolari per modellare in modo efficiente insiemi di dati tabellari di qualsiasi dimensione. XGBoost è un’implementazione di gradient boosting molto veloce e scalabile: i modelli basati su XGBoost vincono regolarmente competizioni online di data science e sono usati su larga scala in vari settori. In questo corso imparerai a usare questa potente libreria insieme a pandas e scikit-learn per costruire e ottimizzare modelli di supervised learning. Lavorerai con insiemi di dati reali per risolvere problemi di classificazione e regressione.

Supervised Learning with scikit-learn

Esplora i fondamenti del gradient boosting, concentrandoti su XGBoost per la regressione e ottimizzazione.

Extreme Gradient Boosting con XGBoost

Impara le basi del gradient boosting e crea modelli di machine learning all'avanguardia usando XGBoost per risolvere problemi di classificazione e regressione.

Limiti della grid search e della random search

Create Your Free Account