Practicality of Hybrid Neural Networks with Time Series Data


I was wondering if anyone could give me an insight into the practicality of using time-series data with a hybrid neural network. In particular, financial market data for an agent to pick a suitable action such as buy or sell.

My main concerns are training time and also finding good hyper-parameters.

I was wondering if anyone has any experience with this and could share their thoughts?

Thanks! James

1 Like

Hey @James_Ellis,

Good question, I would also be curious. The only thing that jumps to my mind is that this demo could be of relevance: If variational models have a Fourier-like structure, then time series data should be an interesting candidate to try.

About training time, the usual warning applies; trainable quantum circuits are still a research tool and cannot compete with long-standing classical techniques.

Good luck!

1 Like