Product Defect Probability Tree

Product defect probability tree

This probability tree separates a product's actual condition from the result of an inspection. A defective item can be found or missed, while a good item can be falsely rejected or pass. The branch probabilities describe how the inspection behaves within each condition. This is a useful structure for quality teams because it distinguishes the true state from the observed result. Add the proportion of defective products at the first split when you need overall probabilities, such as the chance that a rejected item was genuinely defective. The same pattern also applies to tests, alarms, and screening processes. It gives readers a concise reference they can reuse when discussing the result with colleagues.

類型 決策樹標準 Howard-Raiffa / CART-sklearn / taxonomy引擎 schematex-decisiontree更新時間 2026/9/24

需求

“A probability tree for product inspection: Defective or Not defective, then inspection results. A defect is found with probability 0.9, and a good product is falsely rejected with probability 0.05.”

接著試試Add the base rate of defective productsChange defect detection to 0.95

這張圖裡有什麼

看懂背後的決策。

01Actual state

Separate defective from good products.

02Test result

Show found, missed, false reject, and pass.

03Conditional rates

Put inspection accuracy on each branch.

Use for explaining sensitivity and false-positive rates in quality control.

瀏覽所有決策樹範本 →

常見問題

What is a false reject?

A good product that the inspection incorrectly marks as defective.

What is a missed defect?

A defective product that passes the inspection.

Can I add product defect rates?

Yes. Put those probabilities on the first split.