Skip to content

Support for StandardScaler in cuml.accel #7765

@csadorf

Description

@csadorf

Is your feature request related to a problem? Please describe.
StandardScaler is a commonly used preprocessing transformer in scikit-learn pipelines, but currently lacks cuml.accel acceleration support. Users cannot benefit from GPU acceleration when using StandardScaler via cuml.accel.

Describe the solution you'd like
Enable cuml.accel to automatically accelerate sklearn.preprocessing.StandardScaler using cuML's GPU implementation when appropriate hardware and conditions are available.

Describe alternatives you've considered

  • Manually using cuml.preprocessing.StandardScaler directly instead of scikit-learn's version
  • Using cuML's estimators throughout the pipeline rather than relying on cuml.accel

Additional context
StandardScaler is fundamental to many ML workflows and would significantly benefit from GPU acceleration.

Metadata

Metadata

Assignees

Labels

cuml-accelIssues related to cuml.accelfeature requestNew feature or request

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions