AmirhosseinHonardoust/Teaching-Neural-Networks-to-Imagine-Tables

A comprehensive deep dive into how Variational Autoencoders (VAEs) learn to generate realistic synthetic tabular data. This project explores latent space learning, probabilistic modeling, and neural creativity, combining data privacy, interpretability, and generative AI techniques in a structured format.

25
/ 100
Experimental

This project helps data analysts and scientists create realistic, synthetic versions of their structured data tables, such as customer records or financial logs. You provide an existing dataset with mixed data types and numerical columns, and it generates a new table that reflects the original's patterns and relationships without exposing any real individual data. This is ideal for anyone needing to share or work with sensitive tabular information.

Use this if you need to generate new, privacy-preserving tabular datasets that statistically resemble your original data for tasks like sharing with external partners or public use.

Not ideal if your dataset has many columns with a large number of unique categorical values, or if you need to enforce strict logical rules (e.g., 'age cannot be negative') in the generated data.

data-privacy synthetic-data-generation data-augmentation data-simulation analytics
No Package No Dependents
Maintenance 6 / 25
Adoption 6 / 25
Maturity 13 / 25
Community 0 / 25

How are scores calculated?

Stars

22

Forks

Language

License

MIT

Last pushed

Nov 10, 2025

Commits (30d)

0

Get this data via API

curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/AmirhosseinHonardoust/Teaching-Neural-Networks-to-Imagine-Tables"

Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.