Bayesian的评论 · · · · · · · · · · ( 评论2 )
- 你们想看但就是找不到对不对?!
-
Bayesian
(mathematician at heart)
评论:
Heuristics
Bayesian的电影 · · · · · · ( 8部看过 )
1. "Not to be a republican at twenty is proof of want of heart; to be one at thirty is proof of want of head."
---- it still holds if one replaces "republican" with "non-Bayesian".
2. “我们听过贝叶斯定理,却仍旧过不好这一生。”
---- evidence that Bayesian inference is computationally hard
The Fundamental Theorem of Life:
Life is magic, but only in hindsight.
Proof:
The proof is trivial if we apply the Time Trick, we thus omit the details here due to lack of space.
The Time Trick:
Define x(t) as the life state at day t, T as the live-a-day operator such that T(x(t)) = x(t+1), T^N = TT...T as the N-times composition of T. The Time Trick states that there exist integers n1, n2, 0 < n1 < n2 < 10*n1, and some way of living T, such that D(x, T^(n1)(x)) << 1, and D(x, T^(n2)(x)) >> 1, for some subjective semi-metric of life, D.
Footnote:
Empirical results suggest that n1 is approximately 365. Depending on the measure D one choose, n2 could vary but is typically no greater than 2000.
---- it still holds if one replaces "republican" with "non-Bayesian".
2. “我们听过贝叶斯定理,却仍旧过不好这一生。”
---- evidence that Bayesian inference is computationally hard
The Fundamental Theorem of Life:
Life is magic, but only in hindsight.
Proof:
The proof is trivial if we apply the Time Trick, we thus omit the details here due to lack of space.
The Time Trick:
Define x(t) as the life state at day t, T as the live-a-day operator such that T(x(t)) = x(t+1), T^N = TT...T as the N-times composition of T. The Time Trick states that there exist integers n1, n2, 0 < n1 < n2 < 10*n1, and some way of living T, such that D(x, T^(n1)(x)) << 1, and D(x, T^(n2)(x)) >> 1, for some subjective semi-metric of life, D.
Footnote:
Empirical results suggest that n1 is approximately 365. Depending on the measure D one choose, n2 could vary but is typically no greater than 2000.
Bayesian的广播 · · · · · · ( 全部 )
- 未登录无法查看更多内容 -
本页永久链接: https://www.douban.com/people/posterior/
订阅Bayesian的收藏:
feed: rss 2.0