Where does the skill store its outputs?

It follows a strict project structure, saving logic to modular Python files, experiments to notebooks, and processed data to the project's data directory.

How does this skill help with categorical data?

It provides standardized methods for various encoding strategies such as One-Hot Encoding, Label Encoding, and Target Encoding to prepare non-numeric data for modeling.

What libraries does this skill utilize?

The skill is designed to work with standard Python data science libraries including pandas, numpy, and scikit-learn.

Does this skill support domain-specific feature creation?

Absolutely. The framework explicitly includes a step for applying domain knowledge and business logic to ensure features are contextually relevant.

Can this skill handle time-series data?

Yes, it includes specific procedures for creating lag features and moving averages, which are critical for time-dependent datasets.

ML Feature Engineering

Name: ML Feature Engineering
Author: root-5

byroot-5

0•

数据科学与机器学习

Generates and optimizes high-impact features from raw datasets to boost machine learning model performance.

This skill provides a structured framework for the feature engineering phase of the machine learning lifecycle, enabling Claude to systematically transform raw data into predictive signals. It guides the implementation of domain-specific logic, mathematical transformations like Box-Cox and scaling, advanced encoding techniques, and statistical aggregations. By following a standardized workflow, it ensures that feature creation is reproducible and well-documented, making it an essential tool for data scientists using agentic pipelines to accelerate model development.

主要功能

01Feature interaction and statistical aggregation for complex pattern recognition

020 GitHub stars

03Specialized time-series feature generation including lags and moving averages

04Standardized output structure for notebooks, modules, and processed datasets

05Automated data transformations including log, Box-Cox, and normalization

06Comprehensive encoding methods like One-Hot, Label, and Target encoding

使用场景

01Preprocessing datasets for gradient boosting and deep learning models

02Automating the feature discovery process within an agentic ML pipeline

03Improving model accuracy by synthesizing new variables from raw tabular data

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add root-5/agentic-ml-pipeline feature-engineering

For use in Claude.ai and ChatGPT

主要功能

01Feature interaction and statistical aggregation for complex pattern recognition

020 GitHub stars

03Specialized time-series feature generation including lags and moving averages

04Standardized output structure for notebooks, modules, and processed datasets

05Automated data transformations including log, Box-Cox, and normalization

06Comprehensive encoding methods like One-Hot, Label, and Target encoding

使用场景

01Preprocessing datasets for gradient boosting and deep learning models

02Automating the feature discovery process within an agentic ML pipeline

03Improving model accuracy by synthesizing new variables from raw tabular data

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add root-5/agentic-ml-pipeline feature-engineering

For use in Claude.ai and ChatGPT