Study of new designs in quantum neural networks

This repository contains code necessary to reproduce the figures and simulation results of the paper "New designs in quantum neural networks" by Felix Petitzon. Link ArXiv: https://arxiv.org/abs/2203.07872

Compared are:

• classical deep neural network
• quantum neural network separated into feature map and variational circuit
• same but with classical net appended
• a combined qnn, where feature and variational are merged into one fully parameterized circuit
• same but with classical net appended

The classes folder contains all needed classes and utility functions.

The jupyter notebooks fit the models to the data; and produce the plots of the paper.