Automate ML evaluation with baselines, stratified cross-validation, and leakage checks.
npx skills add https://github.com/Yeachan-Heo/My-Jogyo --skill ml-rigor
Please help me install this Agent Skill. Skill: ml-rigor Source: https://github.com/Yeachan-Heo/My-Jogyo/tree/main/skills/ml-rigor Command: npx skills add https://github.com/Yeachan-Heo/My-Jogyo --skill ml-rigor
Enforces baseline comparisons, cross-validation, interpretation, and leakage prevention for ML pipelines.
Load your dataset, define a model, and run the ml-rigor workflow to generate baselines, CV metrics, and interpretation artifacts ready for reporting.
High-intent search queries and answers about installing and using this skill.
Cross-validation prevents overfitting by splitting data into multiple stratified folds, training on some, and validating on others. This skill automates stratified CV, reporting mean and std metrics with confidence intervals for reliable baseline comparisons.
Baseline model comparisons are required to prove a new model adds value over a simple dummy or heuristic model. This skill enforces baseline-driven evaluation, generating comparison reports before you claim any performance improvements.
Permutation importance interprets predictions by measuring the drop in model score when a feature's values are shuffled. This skill provides permutation importance and SHAP analyses to generate interpretable reports explaining feature impact.
You can run stratified cross-validation for classification and regression by defining your model and dataset in the workflow. The skill automates CV across small to large datasets, outputting fold summaries and error analysis artifacts.
Yes, ML model evaluation supports slicing performance by data segments to analyze failure modes. The skill's error analysis capabilities detect leakage and break down metrics across different data segments within a reproducible workflow.
secondskyTrain machine learning models with scikit-learn, PyTorch, and TensorFlow.
hoanghn61Converts TechStackReport JSON into Markdown, HTML, and PDF exports.
LogauaEngstromExplain model predictions by attributing feature contributions with SHAP values.
Snowflake-LabsA concise summary of your roleplay character: name,</name> personality, and/or behavior, formatted as a single string sentence. Typically不超过50字，此处省略
