Training Machine Learning Models

Use this skill when the user wants to fit a model, choose algorithms, and produce a trained artifact.

Overview

This skill owns the training loop: selecting candidate models, fitting them with a validation strategy, and producing reusable model outputs.

When to Use This Skill

• Training a classifier or regressor from prepared data
• Selecting among candidate algorithms and hyperparameters
• Saving a model, pipeline, or reproducible training recipe

Not For / Boundaries

• Metric selection and benchmark comparison alone: use evaluating-machine-learning-models
• Feature creation or transformation strategy alone: use engineering-features-for-machine-learning
• Train/test contamination review: use ml-data-leakage-guard

Typical Outputs

• Training plan and candidate model shortlist