A combinatorial problem related to the classical probability

Jiang Zhou
DOI: https://doi.org/10.1016/j.disc.2024.114067
2024-05-04
Abstract:In the classical probability model, let $f(n)$ be the maximum number of pairwise independent events for the sample space with $n$ sample points. The determination of $f(n)$ is equivalent to the problem of determining the maximum cardinality of specific intersecting families on the set $\{1,2,\ldots,n\}$ . We show that $f(n)\leq n+1$, and $f(n)=n+1$ if there exists a Hadamard matrix of order $n$.
Combinatorics,Probability
What problem does this paper attempt to address?
The paper primarily explores a combinatorial problem related to classical probability models. Specifically, the author defines a function \(f(n)\), which represents the maximum number of pairwise independent events in a sample space with \(n\) sample points. This problem can be further transformed into determining the maximum cardinality of a specific family of intersections. The main contributions of the author include: 1. **Proving** that \(f(n) \leq n + 1\), and that \(f(n) = n + 1\) when there exists a Hadamard matrix of order \(n\). 2. **Defining** \(g(n)\) as the maximum cardinality of a family of intersections satisfying specific conditions, and proving that \(f(n) = g(n) + 1\). 3. **Proposing** that \(g(n) \leq n\), and that \(g(n) = n\) when there exists a Hadamard matrix of order \(n\). 4. **Utilizing** 2-designs and generalized Johnson graphs to obtain a lower bound for \(g(n)\), and illustrating the application of these results through examples. In summary, the paper aims to study the maximum number of pairwise independent events that can be constructed given the size of the sample space, providing theoretical upper and lower bounds as well as exact values in special cases. Additionally, it discusses graph theory concepts related to these results, such as generalized Johnson graphs and the clique number of power set graphs.