Weak type estimates associated to Burkholder's martingale inequality (Q925118): Difference between revisions

Given a probability space \((\Omega, {\mathcal A}, \mu)\), let \({\mathcal A}_1, {\mathcal A}_2, \dots\) be a filtration of \(\sigma\)-subalgebras of \({\mathcal A}\), and let \(E_1, E_2, \dots\) denote the corresponding family of conditional expectations. Given a martingale \(f = (f_1, f_2, \dots)\), adapted to this filtration and bounded in \(L_p(\Omega)\) for some \(2 \leq p < \infty\), Burkholder's inequality claims that \[ \| f \| _p \sim_{c_p} \left\| \left( \sum_{k = 1}^{\infty} E_{k-1}(| df_k| ^2) \right)^{1/2} \right\| _p + \left( \sum_{k = 1}^{\infty} \| df_k \| ^p_p \right)^{1/p}. \] Motivated by quantum probability, \textit{M.\,Junge} and \textit{Q.--H.\thinspace Xu} [Ann.\ Probab.\ 31, No.\,2, 948--995 (2003; Zbl 1041.46050), Bull.\ Lond.\ Math.\ Soc.\ 37, No.\,2, 243--253 (2005; Zbl 1088.46037)] recently extended this result to the range \(1 < p < 2\). In the paper, Burkholder's inequality is studied for \(p = 1\), for which the techniques must be different. Quite surprisingly, two non-equivalent estimates are obtained, which play the role of the weak type \((1, 1)\) analog of Burkholder's inequality. As applications, new properties of Davis' decomposition for martingales are obtained.