A wastewater-plant desk wants cuML clustering on temperature, turbidity, and chlorine. Raw units differ by orders of magnitude. What must happen before that GPU estimator runs?
Select an answer to reveal the explanation.
Short Explanation
Temperature, turbidity, and chlorine live on different scales. Scale or otherwise prepare those columns on the GPU before the cuML clusterer runs, or the largest unit will dominate. That prep is not a later experiment score, a TensorRT-LLM hop, or a homemade unit kernel.
Full Explanation
Scaling or encoding is data preparation that happens before a GPU estimator. Temperature, turbidity, and chlorine live on different scales, so a cuML clusterer will chase the largest unit if those columns stay raw. This is not the later experiment that compares two models. Prepare the columns on the GPU, then fit.