# Data Science

> ML engineering expert for feature engineering, model debugging, and production pipelines

- Skill: `lubluniky/data-science` (Agent Skill, multi-file: 3 files)
- Install (CLI): `npx skillmds@latest add lubluniky/data-science`
- Raw SKILL.md: https://api.skillmd.com/api/skills/lubluniky/data-science/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: lubluniky (https://skillmd.com/u/lubluniky)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/lubluniky/data-science

---


# Data Science Skill

Use this skill for production ML systems, not basic model training:

## When to Use

- **Feature engineering**: Designing features that improve model performance
- **Model debugging**: Understanding why models underperform, identifying data issues
- **Pipeline optimization**: Building efficient training/inference pipelines
- **Experiment tracking**: Reproducible ML experiments, versioning, A/B testing
- **ML in production**: Serving, monitoring, retraining strategies

## Core Capabilities

1. **Feature engineering**: Transform raw data into predictive features, handling categorical data, time series
2. **Model diagnostics**: Identify overfitting, underfitting, data leakage, distribution shift
3. **Pipeline design**: End-to-end ML pipelines with proper train/val/test splits, cross-validation
4. **Production ML**: Model serving, monitoring, performance tracking, retraining triggers
5. **Experiment management**: Track hyperparameters, metrics, reproducibility

## Progressive Disclosure

- [ml-engineering.md](./ml-engineering.md) - Production ML systems, serving, monitoring
- [reinforcement-learning.md](./reinforcement-learning.md) - RL algorithms, environments, training strategies

## Not For

- Basic pandas/numpy operations (already covered)
- Simple model.fit() calls
- Generic data visualization