Tutorial 3

Machine Learning on GPUs

These notebooks apply GPU-native ML workflows to tabular prediction. Participants train and interpret cuML logistic regression, then execute XGBoost with GPU settings, train-test split strategy, model inspection, and prediction quality evaluation using ROC and AUC. A CPU comparison path highlights practical performance differences and trade-offs.

  • cuML logistic regression and explainability-style checks ()
  • Open In Colab
  • GPU XGBoost training/inference/evaluation, optional CPU comparison (https://colab.research.google.com/github/D-Barradas/Accelerated-Data-Science-with-RAPIDS/blob/main/part3/3-06_xgboost.ipynb)
  • Open In Colab

What you will do

  • open a ready-to-run Colab notebook
  • work through a GPU-based machine learning training workflow
  • experiment with tuning and evaluation in an interactive environment
  • adapt the notebook for follow-up research projects

Notebook source


This site uses Just the Docs, a documentation theme for Jekyll.