An Application of Quantum Computing to Modeling Human Decision-Making

We present a quantum cognitive model that can be implemented on a quantum machine. This model uses solutions of the Schrödinger equation to model human cognitive control and decision making in a standard paradigm in psychology: the two-alternative forced-choice (2AFC) task, where participants are forced to choose between two competing alternatives. In the model, the two alternatives map to a multiple square well potential energy function. The depths of the wells correspond to attentional control, while the widths corresponds to features of the stimulus. Using this potential energy, the time-independent Schrodinger equation can be solved for an energy eigenstate basis of the bound states of the system, with higher energies corresponding to higher levels of focus. The full time-dependent Hamiltonian involves an oscillating field that evolves the system from ground state to superpositions of the excited states. The probabilities of the choices in the task are calculated as the probability of finding particles in a particular potential well. Furthermore, the quantum framing of this problem makes this problem suitable for mapping to quantum computing implementations. The state of the system follows the trajectory of a ground state, which can be mapped to a variational quantum eigensolver (VQE) implementation. Meanwhile, the time-dependency and oscillating field can be simulated using current quantum simulation methods and Rabi oscillations. Since our problems can be mapped onto currently-available NISQ prototypes of 2040 qubits, they represent an application domain with tractable near-term quantum-computing resource requirements.