Statistical complexity of quantum circuits

In theoretical machine learning, the statistical complexity is a notion that measures richness of hypothesis space. this work, we apply particular measure complexity, namely, Rademacher to quantum circuit model in computation and study how depends on various parameters. particular, investigate dependence resources, depth, width, number input output registers circuit. To scales with resources circuit, introduce magic resource based $(p,q)$ group norm, which quantifies amount channels associated These dependencies are investigated following two settings: (i) where entire treated as single channel, (ii) each layer separate channel. The bounds obtain can be used constrain capacity neural networks terms their depths widths well network.