When should I use this Claude Code skill?

Use it at the start of any ML project, when a current model is underperforming, or when you need to balance high accuracy with production constraints like low latency and small memory footprints.

Can this skill help with production deployment?

Absolutely. It integrates deployment constraints like inference time and hardware limits into the initial selection process, preventing the choice of models that are too heavy for production.

Does it support deep learning architectures?

Yes, it includes specific heuristics for deep learning in Computer Vision and NLP, while also providing guidance on when simpler models are more effective for smaller or tabular datasets.

How does it handle imbalanced datasets?

The framework provides specific guidance on using resampling, class weights, and robust metrics such as F1-score and AUC-ROC to ensure models don't simply predict the majority class.

What is the ML Model Selection Framework?

It is a systematic methodology designed to help developers and data scientists choose the most appropriate machine learning algorithm based on specific project characteristics, data properties, and constraints.

ML Model Selection Framework

Name: ML Model Selection Framework
Author: lev-os

bylev-os

0•

データサイエンスとML

Guides the selection of optimal machine learning algorithms by analyzing problem structure, data properties, and production constraints.

The ML Model Selection Framework provides a structured methodology for developers and data scientists to move beyond trial-and-error when building AI systems. It systematically evaluates project requirements—such as data size, feature types, interpretability needs, and latency constraints—to recommend the most effective algorithm families, ranging from classical linear models to advanced deep learning architectures. By following this diagnostic approach, users can avoid common pitfalls like over-engineering, ensure model generalizability, and justify architectural decisions with empirical benchmarks and clear success metrics.

主な機能

01Constraint-aware selection balancing accuracy against latency and memory

02Benchmarking workflows for comparing candidate models against baselines

03Data-driven heuristics for matching algorithms to dataset size and type

04Diagnostic problem classification for supervised and unsupervised tasks

050 GitHub stars

06Anti-pattern identification to prevent overfitting and data leakage

ユースケース

01Justifying model choice for highly regulated industries requiring interpretability

02Selecting a baseline model for a new predictive maintenance project

03Optimizing an existing NLP pipeline for production latency requirements

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add lev-os/agents ml-model-selection

For use in Claude.ai and ChatGPT

Download Skill