XGBoost for Classification: Building Accurate and Efficient Models
summary
Shows using XGBoost for binary classification on the UCI wine dataset: splitting data, fitting and predicting with model.fit, and searching hyperparameters via grid, random and cross‑validation. Demonstrates pipelines that handle missing values and parallel training, reports about 80% test accuracy with higher training scores (overfitting risk), uses feature‑importance plots (chlorides, total sulfur dioxide, alcohol) and applies regularization to reduce variance.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 1:05What is classification
- 4:20XGBoost for classification
- 7:33Wine data prep
- 9:16Continuous to binary
- 11:03Train & eval model
- 14:16Metrics & confusion matrix
- 15:43Hyperparameter tuning
- 19:35Grid & Random Search
- 22:02Feature importance
- 24:09Model comparison
- 26:41XGBoost advantages
- 28:45Summary
- 30:10Q&A
related talks
43mDeveloper Accessible Machine Learning
32mDeep Dive into scikit-learn's HistGradientBoosting Classifier..
40mGradient Boosting Machines (GBM): from Zero to Hero (with R and Python code)
38mCatBoost - the new generation of gradient boosting
01h 28mMastering gradient boosting with CatBoost
01h 04mGo with the flow–Track your machine learning lifecycle
37mGradient Boosting for Data with Both Numerical and Text Features
16mStop Failures Before They Happen: Gradient Boosting