GPU Acceleration Demo
This demo shows how to train a model on the forest cover type dataset using GPU acceleration. The forest cover type dataset has 581,012 rows and 54 features, making it time consuming to process. We compare the run-time and accuracy of the GPU and CPU histogram algorithms.
This demo requires the GPU plug-in to be built and installed.
The dataset is automatically loaded via the sklearn script.