🔒 Differential Privacy Budget Calculator

Epsilon & Delta Planner · Composition theorems & noise scaling
1.00
1.0e-6

🧠 What is epsilon? Epsilon bounds how much a single person's data can change the output. ε = 0.1 → very private; ε = 10 → weak privacy. Think of it as "max information leakage" per query.

📈 Privacy Budget Breakdown

Per-query ε
0.010
Total privacy cost (ε_total)
1.00
Privacy-utility tradeoff
0.50
Recommended noise scale
1.00
🔹 Laplace mechanism: scale = 1.00  |  Gaussian (δ): σ = 1.00

Privacy budget depletion

Noise distribution (ε=0.5,1,2)