Training My Very First Variational Quantum Neural Network

Earlier this year, I was lucky enough to take a pretty advanced quantum computing class with an authority in the field. It was low-stakes and very exploratory, which I honestly liked a lot. I walked away mostly feeling curious, which was probably the whole point of the class.

Anyway, onto the fun part.

Below is a notebook where I train my very first variational quantum neural network (QNN) on the classic two half-moons dataset. There’s nothing fancy or state-of-the-art going on here. The goal was just to get something working end-to-end and to build some intuition for how variational circuits actually behave when you treat them like a learning model.