I am learning how to use pennylane to carry portfolio optimization. Kindly, provide links to any tutorials, github etc.
Hi @Aryabhat! Welcome to the PennyLane discussion forum.
You are asking a very broad question: there are many formulations of the portfolio optimization problem, and for each one many algorithms to solve the problem. Could you please elaborate more on what exactly you are interested in?
We don’t currently have any demos aimed specifically at portfolio optimization, but there are several demos discussing applications to optimization problems – see the “Optimization” tab in the research demos page.
Many people use the Quantum Approximate Optimization Algorithm (QAOA) as a versatile approach for solving optimization problems. You can also take a look at our QAOA demo.
Hope that helps!
I am trying to build a hamiltonian for optimizing trade loan portfolio. Unable to frame the ground state Hamiltonian. Please guide me.
Hi @Aryabhat. It sounds like an exciting problem to work on. It’s difficult to help you much though without getting some more details on a specific issue/problem you’re having. The demos that @jmarrazola linked to above could be a good start to explore optimization problems with PennyLane.
I would recommend you to play around a bit with PennyLane, and implement a rough idea of what you want to achieve, and then we could help you out with potential issues or challenges, related to the specific PennyLane implementation, that you might face.
Let us know how it goes!
Hi Theodor, Thanks for your reply. Is there any option for using logistics regression in pennylane. Any work around available. Any work done in this regard.
Hi @Aryabhat, so far PennyLane has focused on providing the fundamental building blocks of quantum differentiable programming. While we don’t have a module focused specifically on logistic regression, if you can write the problem as a composition of differentiable operations, then it should be possible to leverage PennyLane. I’d recommend checking out similar tutorials on using logistic regression in Torch or TensorFlow, and then think about how a quantum circuit can be integrated into the model.