Jayasri summability (Q698583)

The summability of the title is defined by a matrix \(J(f_v, h_v)= (a_{nk})\) with \(a_{00}= 1\), \(a_{0k}= 0\) \((k= 1,2,\dots)\) and \[ \prod^n_{v=1} (f_v(z) h_v+ 1- h_v)= \sum^\infty_{k=0} a_{nk} z^k, \] where \((f_v)\) is a sequence of entire functions and \((h_v)\) is a complex sequence. Let \((a_{nk}(x))\) be given by \[ \prod^n_{v=1} (f_v(z) h_v(x)+ 1- h_v(x))= \sum^\infty_{k=0} a_{nk}(x) z^k, \] where \(f_v(1)= 1\) and \(f^{(k)}_v(0)\geq 0\) (\(v= 1,2,\dots\); \(k= 0,1,\dots\)) and \(h_v\) are functions defined on \([0,1]\) with \(0\leq h_v(x)\leq 1\) (\(v= 1,2,\dots\); \(x\in [0,1]\)). (Note that for \(f_v(z):= z\) and \(h_v(x):= x\) \((v= 1,2,\dots)\), \(J(f_v, h_v)\) is the Euler marix and \(J_n(g)(x)\) is the \(n\)th Bernstein polynomial.) For real functions \(g\) bounded on \([0,\infty)\) and continuous on \([0,1]\), the author defines positive linear operators \(J_n\) by \[ J_n(g)(x):= \sum^\infty_{k=0} a_{nk}(x) g(k/n)\qquad(n= 0,1,\dots) \] and gives three conditions which guarantee the equality \(\lim_n J_n(e_k)(x)= e_k(x)\), where \(e_k(x):= x^k\) (\(k= 0,1,2\); \(x\in [0,1]\)). Then, by the Korovkin theorem, \(\lim_n J_n(g)(x)= g(x)\) for all \(g\) (Theorem 2.1). For example (Corollary 2.3), if \(f_v'(1)= 1\) \((v= 1,2,\dots)\), \((f''_v(1))\in \ell^\infty\) and \((h_v(x))\) is \((C,1)\)-summable to \(x\), then \(J_n(g)\to g\) on \([0,1]\). Quantitative estimates for \((J_n)\) are given and their probabilistic interpretations are discussed.