Higher-order Stirling-number monotonicity and log-concavity conjectures

From papers

Let S(n,k)S(n,k) denote the Stirling numbers of the second kind, and let S1(n,k)\mathfrak{S}_{1}(n,k) be the quantity defined by the paper's equation defining S1(n,k)\mathfrak{S}_{1}(n,k). Recursively define

S+1(n,k)=S2(n,k1)S(n,k2)S(n,k)\mathfrak{S}_{\ell+1}(n,k)=\mathfrak{S}_{\ell}^2(n,k-1)-\mathfrak{S}_{\ell}(n,k-2)\mathfrak{S}_{\ell}(n,k)

and, for nk+2n\geq k\geq \ell+2, define

S(n,k)=S+1(n,k)S(n,k).\mathcal{S}_{\ell}(n,k)=\frac{\mathfrak{S}_{\ell+1}(n,k)}{\mathfrak{S}_{\ell}(n,k)}.

Higher-order Stirling-number conjecture. The following claims are valid: (1) for fixed ellNell\in\mathbb{N} and n+3n\geq \ell+3, the finite sequence {S(n,k)}+1kn\{\mathfrak{S}_{\ell}(n,k)\}_{\ell+1\leq k\leq n} is logarithmically concave in kk; (2) for fixed nk3n\geq k\geq 3, the finite sequence {S(n,k)}1k1\{\mathfrak{S}_{\ell}(n,k)\}_{1\leq \ell\leq k-1} is strictly increasing in \ell; (3) for fixed N\ell\in\mathbb{N} and nk+1n\geq k\geq \ell+1, the sequence {S(n+m,k+m)}m0\{\mathfrak{S}_{\ell}(n+m,k+m)\}_{m\geq 0} is strictly increasing in mm; (4) for fixed N\ell\in\mathbb{N} and k+1k\geq \ell+1, the sequence {S(n,k)}nk\{\mathfrak{S}_{\ell}(n,k)\}_{n\geq k} is strictly increasing in nn; (5) for fixed nk+2n\geq k\geq \ell+2, the sequence {S(n+m,k+m)}m0\{\mathcal{S}_{\ell}(n+m,k+m)\}_{m\geq 0} is strictly increasing in mm; and (6) for fixed k+2k\geq \ell+2, the sequence {S(n,k)}nk\{\mathcal{S}_{\ell}(n,k)\}_{n\geq k} is strictly increasing in nn.

Progress summary

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Sources & referencesView supporting material

Primary source

Feng Qi, “Diagonal recurrence relations, inequalities, and monotonicity related to Stirling numbers”, arXiv:1402.2040 (2014).

Solutions 0

No solutions have been posted yet.