Concrete strength predictor for mix design

The brief

A ready-mix concrete supplier makes and crushes test cubes to check every new mix, which takes 28 days. Build a model that predicts compressive strength from the ingredients and age, and use it to suggest cheaper mixes that still meet the grade. Good looks like: a model with known error, the effect of each ingredient shown clearly, and a small mix-design helper.

What you will build

1. Explore how each ingredient and age affects strength
2. Train regression models and compare errors
3. Explain predictions with partial dependence or SHAP
4. Build a helper that checks a proposed mix
5. Discuss safety margins and when to still test cubes

Free, open data to build it with

• Concrete Compressive Strength ↗ UCI Machine Learning Repository

Check each dataset's licence before you publish your work. Your mentor gives your team its own version of the brief (data, scope or region), so every team's project is its own.