khedekarpratik0337

khedekarpratik0337 / featureiq

Public

A knowledge graph and meta-learning framework for context-aware automated feature engineering

automl data-science feature-engineering knowledge-graph machine-learning
14
8
100% credibility
Found Apr 16, 2026 at 14 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

FeatureIQ is an open-source Python tool that automatically profiles datasets, recommends feature engineering transformations based on data traits, problem type, and algorithm, and builds ready-to-use pipelines with explanations.

How It Works

1
📖 Discover FeatureIQ

You hear about a helpful tool that automatically improves your data for better predictions, saving hours of manual work.

2
🛠️ Get it ready

You add this friendly assistant to your workspace with a quick and easy setup.

3
📊 Load your data

You bring in your table of numbers, categories, or dates, just like opening a spreadsheet.

4
✨ Let it analyze

You share what kind of prediction you're making and your model choice, and it scans everything to suggest perfect improvements.

5
🔄 Apply the magic

It cleans, reshapes, and enhances your data automatically, handling tricky parts like missing values or odd patterns.

6
📋 Check the story

You review simple explanations with reasons and confidence scores for each change it made.

🎉 Win big

Your models now learn faster and perform better, freeing you to focus on insights instead of data tweaks.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 14 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is featureiq?

FeatureIQ automates feature engineering for ML pipelines by profiling your tabular data and recommending sklearn-compatible transforms tailored to your problem type (like binary classification or time series forecasting) and algorithm (XGBoost, logistic regression, etc.). It draws from a curated knowledge graph of rules backed by ML literature citations, with optional meta-learning from OpenML benchmarks for smarter re-ranking. Built in Python, it slots directly into sklearn pipelines and spits out explainable recommendations, validation scores, and ablation studies.

Why is it gaining traction?

Unlike generic AutoML tools, FeatureIQ is context-aware—factoring in data stats, multicollinearity, and algorithm quirks (trees skip scaling, linears get VIF checks)—while staying lightweight and extensible via YAML rules. Developers dig the quick-start API, rich score reports, and per-transform impact analysis that proves value before committing. As a knowledge graph AI for feature engineering, it feels like a github knowledge base copilot, blending ontology rules with LLM-like meta-learning for reproducible gains.

Who should use this?

ML engineers wrangling messy tabular datasets for sklearn models, especially in classification, regression, or anomaly detection where manual transforms eat hours. Teams building knowledge graph tools or automated pipelines will like the ontology for custom rules and RAG-style explanations. Skip if you're deep in images/text or need end-to-end AutoML.

Verdict

Worth pip-installing for prototyping—solid docs, PyPI-ready, and alpha-stable with CI/pre-commit—but at 14 stars and 1.0% credibility, treat as experimental; contribute rules to mature it. Strong start for context-aware feature engineering.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.