Manushree1005

Manushree1005 / synthetic-data-quality-validator

Public

Validate synthetic datasets using ML utility, similarity & privacy risk

37
0
100% credibility
Found Feb 09, 2026 at 33 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

A simple web app that assesses synthetic datasets by scoring their statistical match to real data, usefulness for machine learning tasks, and privacy protection level.

How It Works

1
🔍 Discover the Tool

You hear about a handy checker that tells if fake data matches real data well enough for your projects.

2
📁 Gather Your Files

Collect your real data spreadsheet and the synthetic one you created, both in simple CSV format.

3
☁️ Upload Your Data

Open the app and easily drag your two data files into place to start the comparison.

4
🎯 Choose Key Column

Pick the special column from your data that matters most, like what you're trying to predict.

5
📊 View the Scores

Instantly see three key scores: how similar the data looks, how useful it is for learning, and how safe it keeps privacy.

✅ Get Overall Quality

Receive a single score out of 100 celebrating how great your synthetic data is for real use.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 33 to 37 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 synthetic-data-quality-validator?

This Python tool lets you validate synthetic datasets against real ones via a Streamlit web app. Upload CSV files for real and synthetic data, select a target column, and it spits out scores for statistical similarity, ML utility (by training on synth and testing on real), privacy risk, plus a composite quality score out of 100. It solves the headache of trusting synthetic data for training models without risking privacy leaks or poor performance.

Why is it gaining traction?

In a world of tools to generate synthetic data, this one focuses on validation—checking utility, similarity distributions, and privacy risks in one dashboard. The hook is its dead-simple interface: no complex setup, just pip install requirements and streamlit run app.py for instant results on your datasets. Developers dig the Python-native approach for quick quality checks before ML pipelines.

Who should use this?

AI researchers tweaking GANs or SDV outputs need it to benchmark synthetic data quality. Data scientists handling privacy-sensitive datasets—like healthcare or finance—can validate risk before training. ML students prototyping models will appreciate the fast feedback on utility and similarity.

Verdict

At 33 stars and 1.0% credibility score, it's a raw prototype with basic docs and no tests—fine for local synthetic data validation experiments, but wait for maturity if you're evaluating production workflows. Solid starting point for Python devs needing quick privacy and quality checks.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.