Aleberriz Agent Skills Analytics Ml

ML Modeling

This skill is planned but not yet implemented.

Conventions for applied machine learning in analytics contexts: predicting outcomes, scoring users, forecasting metrics. This is the prescriptive/predictive domain, distinct from experimentation (which establishes causation through randomization).

When complete, this skill will cover:

• Feature engineering patterns for tabular business data
• Model evaluation: choosing the right metric for the business question (precision vs. recall, RMSE vs. MAE, AUC vs. F1)
• Validation strategy: time-based splits for temporal data, stratification, cross-validation
• Overfitting discipline: regularization, feature selection, holdout discipline
• Forecasting conventions: time series decomposition, seasonality, trend
• Deployment considerations: batch vs. real-time, monitoring, drift detection

References

• Anthropic Agent Skills: check for updated patterns and templates
• scikit-learn: the default ML library for tabular analytics work
• statsmodels: statistical modeling in Python

Source: aleberriz/agent-skills — distributed by TomeVault.