principle of inclusion
When the number of cases of an event is difficult to count directly, it is solved by dividing it into smaller and simpler counts problem partitioning. When the partial events are exclusive, they simply add up, but even if they do not, the inclusion-division principle can be used if the common part can be found. $ | A \cup B | = |A| + |B| - |A \cap B|
https://gyazo.com/afe3d4832a4c5543a2d318ddc95e32d7
When you want to find 000 but it is difficult to find it directly, calculate *** - (1** + *1* + **1) + (11* + 1*1 + *11) - (111)
In particular, if the size f(k) of the set is determined by the number k of 1s
$ \sum_k -1^k \binom{n}{k} f(k)
$ | \bigcap_{i \in V}A_i | = \sum_{S \subseteq V} (-1)^{|S|} | \bigcap_{i\in S} A_i^{\mathsf c} |
$ | A \cup B | = |A| + |B| - |A \cap B|
$ | A^{\mathsf c} \cap B^{\mathsf c} | = |U| - |A| - |B| + |A \cap B|
Name the events properly and transform the equation; be aware of which events can be easily calculated.
Consider the case where N=3 this time. Let #() denote the number of elements in the event. Pi = (the event that Ai=Bi), we want to find #(¬P1 ∧ ¬P2 ∧ ¬P3), and #(P1) or #(P1 ∧ P2) can be easily computed. #(¬P1 ∧ ¬P2) is not so easy to deal with, so it is easy to use de Morgan to convert it to #(¬(P1 ∨ P2 ∨ P3)) and think of it as the inclusive divisor principle. If small, full search
Symmetry results in combination
O(2^N) becomes O(N)
O(3^N) when using the inclusion principle for each subset
This is a heavy I
relevance
---
This page is auto-translated from /nishio/包除原理 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.