Skip to content

Instantly share code, notes, and snippets.

@zackmdavis
Created December 24, 2015 00:41
Show Gist options
  • Save zackmdavis/90c4440ad32fac9a05fb to your computer and use it in GitHub Desktop.
Save zackmdavis/90c4440ad32fac9a05fb to your computer and use it in GitHub Desktop.
Welcome to Mezzanine v. 0.2.0! Privately think of a criterion concerning
natural numbers not greater than 50. This program will attempt to efficiently
infer the nature of the criterion by asking you whether specific numbers do or
do not have the property of satisfying the criterion.
Size of hypothesis space: 21984
This program's belief distribution (over 21984 remaining hypotheses) has an
entropy of 14.424 bits. Learning whether 18 has the property is expected to
reduce the entropy by 1.000 bits.
Does 18 have the property? [Y/n] >> n
On the question of whether 18 has the property, you said false.
This program's belief distribution (over 10995 remaining hypotheses) has an
entropy of 13.425 bits. Learning whether 40 has the property is expected to
reduce the entropy by 1.000 bits.
Does 40 have the property? [Y/n] >> y
On the question of whether 40 has the property, you said true.
This program's belief distribution (over 5432 remaining hypotheses) has an
entropy of 12.407 bits. Learning whether 43 has the property is expected to
reduce the entropy by 0.999 bits.
Does 43 have the property? [Y/n] >> n
On the question of whether 43 has the property, you said false.
This program's belief distribution (over 2638 remaining hypotheses) has an
entropy of 11.365 bits. Learning whether 35 has the property is expected to
reduce the entropy by 1.000 bits.
Does 35 have the property? [Y/n] >> y
On the question of whether 35 has the property, you said true.
This program's belief distribution (over 1342 remaining hypotheses) has an
entropy of 10.390 bits. Learning whether 20 has the property is expected to
reduce the entropy by 1.000 bits.
Does 20 have the property? [Y/n] >> y
On the question of whether 20 has the property, you said true.
This program's belief distribution (over 687 remaining hypotheses) has an
entropy of 9.424 bits. Learning whether 27 has the property is expected to
reduce the entropy by 0.999 bits.
Does 27 have the property? [Y/n] >> y
On the question of whether 27 has the property, you said true.
This program's belief distribution (over 329 remaining hypotheses) has an
entropy of 8.362 bits. Learning whether 39 has the property is expected to
reduce the entropy by 1.000 bits.
Does 39 have the property? [Y/n] >> n
On the question of whether 39 has the property, you said false.
This program's belief distribution (over 163 remaining hypotheses) has an
entropy of 7.349 bits. Learning whether 22 has the property is expected to
reduce the entropy by 0.988 bits.
Does 22 have the property? [Y/n] >> y
On the question of whether 22 has the property, you said true.
This program's belief distribution (over 71 remaining hypotheses) has an
entropy of 6.150 bits. Learning whether 5 has the property is expected to
reduce the entropy by 0.983 bits.
Does 5 have the property? [Y/n] >> y
On the question of whether 5 has the property, you said true.
This program's belief distribution (over 30 remaining hypotheses) has an
entropy of 4.907 bits. Learning whether 33 has the property is expected to
reduce the entropy by 1.000 bits.
Does 33 have the property? [Y/n] >> n
On the question of whether 33 has the property, you said false.
This program's belief distribution (over 15 remaining hypotheses) has an
entropy of 3.907 bits. Learning whether 29 has the property is expected to
reduce the entropy by 0.971 bits.
Does 29 have the property? [Y/n] >> y
On the question of whether 29 has the property, you said true.
This program's belief distribution (over 9 remaining hypotheses) has an
entropy of 3.170 bits. Learning whether 19 has the property is expected to
reduce the entropy by 0.918 bits.
Does 19 have the property? [Y/n] >> n
On the question of whether 19 has the property, you said false.
This program's belief distribution (over 6 remaining hypotheses) has an
entropy of 2.585 bits. Learning whether 21 has the property is expected to
reduce the entropy by 1.000 bits.
Does 21 have the property? [Y/n] >> y
On the question of whether 21 has the property, you said true.
This program's belief distribution (over 3 remaining hypotheses) has an
entropy of 1.585 bits. Learning whether 31 has the property is expected to
reduce the entropy by 0.918 bits.
Does 31 have the property? [Y/n] >> n
On the question of whether 31 has the property, you said false.
This program infers that a natural number has the property iff it is divisible
by 5 or it is not less than 20 and it is not greater than 29.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment