关于「边做边学:汤普森采样与自适应实验的估计偏差」,《算法营销与归因》给出的定量结论是什么?
结论:汤普森采样按「是最好版本的后验概率」分配流量,在虚构的三创意例子中把后悔从 A/B/C 测试的 33.9 次降到 26.1 次,但选对最好创意的比例从 83% 降到 79%。在两个版本完全相同时,汤普森分配使样本转化率平均偏低 0.33 个百分点,名义置信区间的覆盖率降到 82.5%;用已知分配概率做逆倾向加权可以消除偏差。作者框架中把赌博机写成「优化转化」的工具是对的,但不能把它当作测量效果的工具;两者的目标不同,应当在实验设计时就说清楚要哪一个。
数据口径:一笔成交该算在哪个广告渠道头上?一个 A/B 测试什么时候可以停?一位客户还值多少钱?本书是教学与研究用的数学模型,不是营销、投资或法律建议。 第八章给出归因、实验、生命周期价值与预算四步核查表,以及一项预注册的模拟基准研究:在独立第三方冻结、已知真实增量的合成多渠道数据生成器上,比较观测沙普利归因与末次点击对真实增量份额的误差(唯一主要结局为二者平均绝对误差之差,最小关注效应 2 个百分点,90% 把握度需要 79 个合成世界、计划 90 个,判定规则写成代码)。
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: 26.1, 33.9, 79, 83
| rates (0.04, 0.045, 0.05), 10,000 visitors, 100 replications | |
| policy regret (conv.) traffic on best final pick correct | |
| A/B/C 6,000 then best | 33.9 (sd 7.2) 52.4% 83% |
| epsilon-greedy 10 % | 25.8 (sd 22.6) 60.7% 74% |
| Thompson sampling | 26.1 (sd 14.6) 60.5% 79% |
| Thompson: visitors shown the 4.0 % creative, median 1037, min 122, max 5496 | |
| always the best | 0 regret; uniform random: 50.0 |
查看完整验算程序
# Three ad creatives with FICTIONAL conversion rates 4.0 %, 4.5 % and 5.0 %; 10,000
# visitors. Policies: (1) A/B/C test with equal split for the first 6,000 visitors,
# then the observed best for the rest; (2) epsilon-greedy (10 %); (3) Thompson sampling
# with Beta(1, 1) priors (Thompson 1933). Reported over 100 replications: expected
# regret (conversions lost against always showing the best creative), share of
# traffic on the best arm, and how often the final pick is the best arm.
import random, statistics
RATES = (0.040, 0.045, 0.050); N = 10_000; REPS = 100
def run(rng, policy):
s, f = [0, 0, 0], [0, 0, 0]; regret, best = 0.0, 0
for t in range(N):
if policy == "ab":
k = t % 3 if t < 6000 else max(range(3), key=lambda i: s[i] / (s[i] + f[i]))
elif policy == "eps":
if rng.random() < 0.10 or t < 30:
k = rng.randrange(3)
else:
k = max(range(3), key=lambda i: (s[i] + 1) / (s[i] + f[i] + 2))
else:
k = max(range(3), key=lambda i: rng.betavariate(s[i] + 1, f[i] + 1))
if rng.random() < RATES[k]:
s[k] += 1
else:
f[k] += 1
regret += RATES[2] - RATES[k]; best += k == 2
pick = max(range(3), key=lambda i: (s[i] + 1) / (s[i] + f[i] + 2))
return regret, best / N, pick == 2, [s[i] + f[i] for i in range(3)]
if __name__ == "__main__":
print(f"rates {RATES}, {N:,} visitors, {REPS} replications")
print("policy regret (conv.) traffic on best final pick correct")
for pol, lab in (("ab", "A/B/C 6,000 then best"), ("eps", "epsilon-greedy 10 %"),
("ts", "Thompson sampling")):
rng = random.Random(1933)
R = [run(rng, pol) for _ in range(REPS)]
reg = [r[0] for r in R]
mr, sr = statistics.fmean(reg), statistics.stdev(reg)
print(f" {lab:22s} {mr:6.1f} (sd {sr:4.1f})"
f" {statistics.fmean(r[1] for r in R):6.1%}"
f" {sum(r[2] for r in R) / REPS:5.0%}")
if pol == "ts":
n_worst = sorted(r[3][0] for r in R)
print(f" Thompson: visitors shown the 4.0 % creative,"
f" median {n_worst[REPS // 2]}"
f", min {n_worst[0]}, max {n_worst[-1]}")
unif = N * (RATES[2] - sum(RATES) / 3)
print(f"always the best: 0 regret; uniform random: {unif:.1f}")唯一数据源与引用
范雅福(Fan, Y.). (2026). 《算法营销、多触点沙普利归因与获客漏斗的随机控制》第 4 章「边做边学:汤普森采样与自适应实验的估计偏差」(版本 1.0.0). 5D Library. https://doi.org/10.5281/zenodo.23836264
DOI: 10.5281/zenodo.23836264 · 阅读本章所在专著 · 5DL-B-048 · v1.0.0 · CC-BY-NC-ND-4.0
BibTeX
@book{fan2026048,
author = {Fan, Yafu},
title = {Algorithmic Marketing, Multi-Touch Shapley Attribution and Stochastic Control of Acquisition Funnels},
year = {2026},
publisher = {5D Library},
edition = {1.0.0},
doi = {10.5281/zenodo.23836264},
url = {https://doi.org/10.5281/zenodo.23836264}
}