On sharp bounds for marginal densities of product measures (Q503274)

Let \(f(x_1,\dots,x_n)=\prod_{i=1}^n f_i(x_i)\) be a probability density on \(\mathbb R^n\), where \(f_i(x_i)\leq 1\). Let \(E\) be a \(k\)-dimensional subspace of \(\mathbb R^n\). Then, the marginal density \(\pi_E(f)\) of \(f\) on \(E\) admits the universal bound \(2^{k/2}\). The main theorem provides a more detailed local bound adapted to the \(f_i\).