How to find the corresponding quantum neural network for a classical neural network (pictured above). How to use the Dirac symbol to represent the mathematical model of the classical neural network, and how to convert the CNN model in the figure above into a quantum circuit? There is a problem that if the loss function uses least squares regression in CNN, how to construct it with circuit?
Hi @RX1, these are very interesting questions.
Why do you need to convert this CNN to a QNN? Often the need to include quantum in our neural networks is motivated by the problem at hand. We may include a quantum circuit as a node in the classical neural network, or often rethink the structure completely. Here you will find an intro into Quantum Convolutional Neural Networks which might be helpful.
If you want to learn more about quantum neural networks I suggest you try out a Demo, such as the one on learning to learn with quantum neural networks.
Finally, regarding your question about the least squares loss function, in this blog post about how to learn quantum machine learning you can find an example. I also suggest reading the full post because it might clarify a lot of questions.
Let me know if this helps and don’t hesitate to come back if you have any further questions!