scientific article; zbMATH DE number 2163299

From MaRDI portal
Publication:4672143





For any fixed integer \(k\geq 2\) and real \(p\), \(0< p< 1\), the authors consider the following particular case of weighted \(k\)-generalized Fibonacci numbers, denoted by \({F_n}^{(k,p)}\), which they call \((k, p)\)-generalized Fibonacci numbers: \({F_n}^{(k,p)}= 0\) for \(0\leq n\leq k-2\), \({F_{k-1}}^{(k,p)}= p/(1- p)\), and \({F_n}^{(k,p)}= [(1- p)/p]({F_{n-1}}^{(k,p)}+ {F_{n-2}}^{(k,p)}+ {F_{n-k}}^{(k,p)})\) for \(n> k\). They employ probabilistic arguments, involving success runs, the total probability theorem and the inclusion-exclusion principle, to establish the following generalization of a result of \textit{D. E. Ferguson} [Fibonacci Q. 4, 270--273 (1966; Zbl 0139.26703)]: \[ {F_n}^{(k,p)}= \sum^{[(n- k+ 1)/(k+ 1)]}_{r=0} (-1)^r p^{k(r+1)-n}(1- p)^{r-1} \left[(1- p){n-k(r+ 1)\choose r}+ {n-k(r+ 1)\choose r-1}\right], \] for \(n\geq k\).











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4672143)