关于「外生与内生:幂律记忆的三类衰减与指数核的半衰期」,《评论裂变分支比》给出的定量结论是什么?

结论:第二章在幂律记忆下复现克兰与索内特的衰减分类,在指数记忆下给出半衰期,n 为 0.95 时从裸核的 6.93 分钟拉长到 138.6 分钟,并用峰值占比区分推起来的峰(中位数 0.538)与长起来的峰(0.093)。

数据口径:书中的帖子、账号、流量、成本与评论价值均为示意参数;全部计算由书中程序生成;本书是非个人化经营建议,只讨论真实用户的评论与平台官方工具买来的曝光,不涉及任何制造评论或虚构稀缺的手段。 第四章用小崎递推、补偿子恒等式与黄金分割求最大似然估计,示意帖的分支比 95 % 区间从 0.42 到 0.92,两小时估计的标准差约 0.186。

English version

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: 0.95, 138.6, 6.93

bare kernel half-life ln2/beta = 6.93 min
n1/(1-n)follow-upshalf-lifet90 (min)rate at 60 min
0.301.430.439.932.90.0004
0.502.001.0013.946.10.0025
0.602.501.5017.357.60.0054
0.805.004.0034.7115.10.0241
0.9010.009.0069.3230.30.0494
0.9520.0019.00138.6460.50.0704
observed half-life 34.7 min with beta 0.1 -> n = 0.800
a long half-life is a near-critical cascade, not a long push.
查看完整验算程序
# Chapter 2. p22: Exponential memory: the decay rate itself carries n.
# Model: Hawkes kernel n beta exp(-beta t). After one outside shock the
# mean excess comment rate is h(t) = n beta exp(-beta (1 - n) t), so the
# excess halves every ln 2 / (beta (1 - n)) minutes and the shock brings
# n / (1 - n) follow-up comments in all. beta = 0.1 per minute (a mean
# reply delay of 10 minutes) is illustrative.
import math

BETA = 0.1
print("bare kernel half-life ln2/beta = %.2f min" % (math.log(2) / BETA))
print("n      1/(1-n)  follow-ups  half-life  t90 (min)  rate at 60 min")
for n in (0.3, 0.5, 0.6, 0.8, 0.9, 0.95):
    k = BETA * (1.0 - n)
    half = math.log(2) / k
    t90 = math.log(10) / k
    h60 = n * BETA * math.exp(-k * 60.0)
    print("%.2f   %6.2f   %8.2f    %7.1f    %7.1f     %.4f"
          % (n, 1 / (1 - n), n / (1 - n), half, t90, h60))
n_obs = 1.0 - math.log(2) / (BETA * 34.7)
print("observed half-life 34.7 min with beta 0.1 -> n = %.3f" % n_obs)
print("a long half-life is a near-critical cascade, not a long push.")

唯一数据源与引用

范雅福(Fan, Y.). (2026). 《霍克斯自激点过程、评论裂变分支比临界值与直播跟风抢购余震半衰期母典》第 2 章「外生与内生:幂律记忆的三类衰减与指数核的半衰期」(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-234 · v1.0.0 · CC-BY-NC-ND-4.0

BibTeX
@book{fan2026234,
  author    = {Fan, Yafu},
  title     = {Hawkes Self-Exciting Point Processes: Branching-Ratio Criticality in Comment Cascades and the Aftershock Half-Life of Livestream Herd Buying},
  year      = {2026},
  publisher = {5D Library},
  edition   = {1.0.0},
  note      = {DOI pending}
}

本书其他章节结论