关于「ε 取多少:2020 年美国人口普查之争与小数字的代价」,《差分隐私与经营统计发布》给出的定量结论是什么?
结论:第六章用后验上限与相对误差两张表读 ε 的取值,并陈美国 2020 年人口普查选区重划数据总预算 19.61 引发的争论:计数为 3 时 ε 为 1 的相对半宽是 0.999。
数据口径:书中参数为示意;全部计算由书中程序生成;本书是非个人化的法律与经营建议,差分隐私是数学保证,不替代法律合规,法规只述大意,以上架前人工复核为准。 第三章处理敏感度:营收按人截断,示意数据中均方根误差 6545 元,占营收 4.09 %;客群分布按并行组合只花一次预算。
对照数据表
下表数字由本章 Python 程序原样打印,可重新运行复现。 程序只用 Python 标准库与结论吻合的打印值: 19.61
| largest posterior after an eps-DP release, by prior belief | |||||
| eps | e**eps | p0 0.001 | p0 0.01 | p0 0.1 | p0 0.5 |
| 0.10 | 1.1 | 0.0011 | 0.0110 | 0.1094 | 0.5250 |
| 0.50 | 1.6 | 0.0016 | 0.0164 | 0.1548 | 0.6225 |
| 1.00 | 2.7 | 0.0027 | 0.0267 | 0.2320 | 0.7311 |
| 2.00 | 7.4 | 0.0073 | 0.0695 | 0.4509 | 0.8808 |
| 2.47 | 11.8 | 0.0117 | 0.1067 | 0.5678 | 0.9220 |
| 5.00 | 148.4 | 0.1293 | 0.5999 | 0.9428 | 0.9933 |
| 8.00 | 2981.0 | 0.7490 | 0.9679 | 0.9970 | 0.9997 |
| 12.20 | 198789.2 | 0.9950 | 0.9995 | 1.0000 | 1.0000 |
| 17.14 | 27784809.3 | 1.0000 | 1.0000 | 1.0000 | 1.0000 |
| 19.61 | 328484430.8 | 1.0000 | 1.0000 | 1.0000 | 1.0000 |
| eps at which prior 0.001 can reach even odds: 6.91 | |||||
| eps at which prior 0.01 can reach even odds: 4.60 |
查看完整验算程序
# Chapter 6. p61: what a given eps allows an attacker to believe.
# Worst case: the attacker knows everyone else and holds prior belief
# P0 that a fact about the target is true. After any eps-DP release the
# posterior is at most P0 * e**eps / (1 - P0 + P0 * e**eps). The rows
# include the totals the U.S. Census Bureau announced for its 2020
# redistricting data (12.2 in the April 2021 demonstration; 17.14 for
# persons, 2.47 for housing units and 19.61 in all, June 2021). This is
# an upper bound for the strongest attacker, not a measured attack rate.
import math
ROWS = [0.1, 0.5, 1.0, 2.0, 2.47, 5.0, 8.0, 12.2, 17.14, 19.61]
PRIORS = [0.001, 0.01, 0.1, 0.5]
def posterior(p0, eps):
up = p0 * math.exp(eps)
return up / (1.0 - p0 + up)
print("largest posterior after an eps-DP release, by prior belief")
print("%7s %13s" % ("eps", "e**eps") + "".join(
"%10s" % ("p0 %g" % p) for p in PRIORS))
for e in ROWS:
print("%7.2f %13.1f" % (e, math.exp(e)) + "".join(
"%10.4f" % posterior(p, e) for p in PRIORS))
for p in PRIORS[:2]:
need = math.log((1.0 - p) / p)
print("eps at which prior %g can reach even odds: %.2f" % (p, need))唯一数据源与引用
范雅福(Fan, Y.). (2026). 《差分隐私拉普拉斯机制与隐私预算会计、公开经营统计而不泄露个人的发布母典》第 6 章「ε 取多少:2020 年美国人口普查之争与小数字的代价」(版本 1.0.0). 5D Library. DOI 待发布
DOI 待发布 · 专著页正式发布后上线(本站暂无专著页) · 5DL-B-296 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026296,
author = {Fan, Yafu},
title = {Differential Privacy, the Laplace Mechanism and Privacy-Budget Accounting: Publishing Business Statistics without Exposing Any Individual},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
note = {DOI pending}
}