Chip yield statistics

The brief

A semiconductor fab sees yield vary from lot to lot. Use process measurements and pass/fail results to compare yield across periods, test which differences are real, and find the process signals most linked to failures. Good looks like: correct tests with multiple-comparison care, effect sizes, and a short list of signals for engineers to check.

What you will build

1. Explore the process signals and outcomes
2. Compare yield across periods with hypothesis tests
3. Correct for testing many signals at once
4. Rank signals linked to failures
5. Write a short report for process engineers

Free, open data to build it with

• SECOM semiconductor manufacturing ↗ 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.