| （英） |
Sparse Identification of Nonlinear Dynamics (SINDy) is a data-driven method for estimation and prediction of nonlinear dynamical systems from time-series data, where nonlinear dynamical systems are represented as a linear combination of apredefined set of basis functions and their coefficients are sparsely estimated from the time-series data. With the development ofquantum information science, quantum circuit learning has been proposed as a quantum-classical hybrid machine learning method,which are assumed to be implemented by NISQ devices without quantum error correction. In this study, we propose SparseIdentification of Quantum Dynamics (SIQDy) that employs quantum circuit learning, where quantum dynamics is expressed asa product of basis quantum circuits, and the associated circuit parameters are sparsely estimated from the time-series data. Assimple examples, we use SIQDy to estimate the quantum dynamics of single-spin and two-spin systems. |