Impulse Ml

Turn Impulse measurement recordings into a machine-learning feature matrix. Use when the user wants to "make ML features from measurement data", "train a model on containers/recordings", build a per-drive / per-recording feature vector from sensor statistics, or feed Impulse output to MLflow or AutoML. Covers computing event-scoped statistics with a sinkless Report, reading them from `report.aggregation_dfs`, and pivoting to one row per container.

databrickslabs Updated 73 repo stars
Impulse architecture

File contents

databrickslabs/impulse commit efb30977de

Frequently asked questions

npx get-skill add databrickslabs/impulse-ml