README.md

LEGO Brick Classification (Classical AutoML)

Model repo: Iris314/classical-automl-model
Dataset: aedupuga/lego-sizes

About

This model classifies LEGO pieces into three types: Standard, Flat, Sloped
Inputs: Length, Height, Width, Studs.

• Model: AutoML (best model = LightGBM, via AutoGluon)
• Task: Tabular classification
• Limitations: Synthetic dataset (augmented from 30 → 300 samples); educational use only.

Demo

Gradio interface allows:

• Single-record input via sliders
• Example inputs (3 presets)
• Prediction (most likely brick type)
• Class probabilities

Acknowledgments

• Model by Xinxuan Tang (CMU), dataset curated by Anuhya Edupuganti (CMU).
• This Gradio app was implemented for CMU course HW3.
• GenAI disclosure: ChatGPT assisted in generating scaffolding code and documentation. All outputs were reviewed and edited.