A reduced order method for nonlinear parameterized partial differential equations using dynamic mode decomposition coupled with k-nearest-neighbors regression

•We propose a novel parameterized dynamic model decomposition (DMD).•DMD is parameterized by coupling with the KNN regression.•The effectiveness of KNN-DMD is validated on several benchmark problems.•KNN-DMD has a comparable performance with POD-ANN on interpolation.•KNN-DMD outperforms POD-ANN on extrapolation.