Portfolio Optimization on a Quantum computer

Qiskit implementation of a Portfolio Optimization.

Requirements

• Python 3.8+
• Qiskit
• CVXPY
• MOSEK
• tqdm

pip install -r requirements.txt

VQE Solver

The VQE Solver takes a Covariance matrix and changes it to an Ising problem. The Ising model's hamiltonian is then minimized with the VQE.

from src.solver.vqe.vqe_solver import VQESolver

from qiskit import BasicAer
from qiskit.utils import QuantumInstance
from qiskit.circuit.library import TwoLocal
from qiskit.algorithms.optimizers import SLSQP

vqe = VQESolver()
vqe.qp(Cov = Cov)
vqe.to_ising(Nq = Nq)

# Prepare QuantumInstance
qi = QuantumInstance(BasicAer.get_backend('statevector_simulator'), seed_transpiler=seed, seed_simulator=seed)

# Select the VQE parameters
N = Cov.shape[0]
ansatz = TwoLocal(num_qubits=N*Nq, 
                    rotation_blocks=['ry','rz'], 
                    entanglement_blocks='cz',
                    reps=args.reps,
                    entanglement='full')
slsqp = SLSQP(maxiter=args.maxiter)
vqe.vqe_instance(ansatz=ansatz, optimizer=slsqp, quantum_instance=qi)

res_vqe = vqe.solve()
print(res_vqe)

License

Apache License 2.0