QAOA Qiskit: Quantum Solution To Classical Problem
Examine QAOA Qiskit for quantum optimisation. Learn the workflow, from building up the cost Hamiltonian and ansatz circuit to solving problems with Sampler and Estimator. Quantum Approximate Optimisation (QAOA) is a popular hybrid quantum-classical method for combinatorial optimisation. A quantum circuit processes quantum states and a classical optimisation mechanism fine-tunes parameters to solve computationally tough problems for traditional computers.
Core Idea and VQE QAOA is a popular method for addressing difficult problems. Its core is an augmentation of the Variational Quantum Eigensolver's optimisation framework. A major difference is that QAOA employs its own “fine-tuned” ansatz, while VQE allows many quantum circuit designs. The ground state of a problem-specific Hamiltonian is the optimum solution to the initial optimisation problem, hence QAOA aims to approach it.
Encoding Problem Coding the classical optimisation problem into a Hamiltonian is essential to employing QAOA. This Hamiltonian appropriately represents the problem's cost function. Challenges like the Max-Cut problem maximise edges linking two groups of vertices in a graph. First, classical problems are explained by minimising a binary variable function (0 or 1 for each node). This creates a Quadratic Unconstrained Binary Optimisation (QUBO) problem. Pauli Z matrices replace binary variables in the QUBO formulation to bridge it to the quantum domain. The cost function Hamiltonian results from this transformation. This Hamiltonian's zero linear terms may simplify Max-Cut scenarios. Finding this cost Hamiltonian's ground state, or lowest energy state, yields the best solution to the initial classical issue. Ansatz Structure QAOA's quantum circuit, or ansatz, is defined by the integer parameter reps. This parameter controls how well the technique approximates the ideal solution and the ansatz's depth or number of layers. Problem and mixer Hamiltonians are alternatively overlaid in the QAOA ansatz. Problem Hamiltonian layers: Phase gates depending on the problem's cost function embed its details into the quantum state. Layers of mixer Hamiltonians: Global X rotations are prevalent in mixer Hamiltonian layers. Superposition is meant to let the quantum circuit explore the diversified solution space. A custom mixer Hamiltonian can be provided for constrained optimisation circumstances when the mixer limits quantum state generation to a feasible subspace. Traditional optimisable factors govern these alternating layers, called gamma and beta angles. Initial_point can be used to set these parameters' initial values during optimisation. Implementing QAOA Qiskit and Workflow Qiskit, IBM's open-source quantum computing framework, provides the QAOA class in qiskit_algorithms. Users can define and optimise quantum circuits with this class. The optimiser (classical optimiser), mixer (for a custom mixer Hamiltonian), initial_state (an optional initial quantum circuit), reps (the integer p parameter for ansatz depth), and initial_point (for initial parameter values) are important when initialising a QAOA object Qiskit's fundamental methodology for QAOA optimisation involves several steps: Classical Inputs to Quantum Problem Mapping The initial stage is to transfer classical inputs to a quantum problem by converting a graph into a QUBO problem, operator cost Hamiltonian, and ansatz parameterised quantum circuit. Problem optimisation for quantum hardware Transpilation optimises abstract quantum circuits for quantum processing units. Transpilation involves mapping qubits, unrolling instructions to the native hardware, routing interacting qubits, and implementing error suppression techniques. Employ Qiskit Primitives Qiskit Runtime primitives like Sampler (which calculates bitstring probability distributions) and Estimator (which computes cost function expectation values) are used in the iterative optimisation loop. A cost function wrapper for the estimator is often used with SciPy's minimise function. After-process and Return After determining the optimal parameters, a sampler runs the quantum circuit one last time. After analysing the bitstring distribution, the greatest probability or lowest cost bitstring is picked as the original problem's solution. Outlook, Apps QAOA can optimise portfolios and Max-Cut. Noise prevents quantum computers from outperforming classical machines in combinatorial optimisation, but algorithm and hardware advances are promising. Researchers are trying quantum heuristics like QAOA on increasingly large problems with more powerful quantum devices.
