Dataset, Code, and Trained Models: Deep Learning-Accelerated Inverse Design of a Nonlinear Duffing-Type Metamaterial
This dataset contains the 25,000-sample training dataset used in "Deep Learning-Accelerated Inverse Design of a Nonlinear Duffing-Type Metamaterial: Amplitude-Programmable, Bifurcation-Aware Bandgap Prediction." Each row corresponds to a sampled design point (nonlinear stiffness k3, mass ratio m2/m1, excitation amplitu...