Nonunitary quantum machine learning

Abstract

We introduce several probabilistic quantum algorithms that overcome the normal unitary restrictions in quantum machine learning by leveraging the linear combination of unitaries (LCU) method. We cover three distinct topics, beginning with quantum native implementations of residual networks (ResNets). We demonstrate that while residual connections between layers of a variational Ansatz can prevent barren plateaus in models, this approach is accompanied by a trade-off in success probability. Second, we implement a quantum analogue of average-pooling layers from convolutional networks using single-qubit-controlled basic arithmetic operators and show that the LCU success probability remains stab..