Skip to main content

Table 1 Leave-one-out cross-validation results

From: Deep learning-based wave digital modeling of rate-dependent hysteretic nonlinearities for virtual analog applications

 

\(1 \textrm{Hz}\)  

\(2 \textrm{Hz}\)

\(5 \textrm{Hz}\)

\(10 \textrm{Hz}\)

\(20 \textrm{Hz}\)

\(50 \textrm{Hz}\)

\(100 \textrm{Hz}\)

\(200 \textrm{Hz}\)

\(500 \textrm{Hz}\)

\(1000 \textrm{Hz}\)

\(\textrm{Average}\)

\(\mathcal {E}(b, \hat{b})\)

\(1.66 \cdot 10^{-4}\)

\(4.10\cdot 10^{-5}\)

\(3.67\cdot 10^{-5}\)

\(2.24\cdot 10^{-5}\)

\(2.45\cdot 10^{-5}\)

\(1.55\cdot 10^{-5}\)

\(3.73\cdot 10^{-4}\)

\(1.05\cdot 10^{-4}\)

\(2.90\cdot 10^{-4}\)

\(4.73\cdot 10^{-4}\)

\(1.55\cdot 10^{-4}\)

\(\mathcal {E}(\mathscr {F}, \hat{\mathscr {F}})\)

\(5.96 \cdot 10^{-4}\)

\(1.46\cdot 10^{-4}\)

\(1.31\cdot 10^{-4}\)

\(7.99\cdot 10^{-5}\)

\(8.75\cdot 10^{-5}\)

\(5.60\cdot 10^{-5}\)

\(1.39\cdot 10^{-3}\)

\(3.90\cdot 10^{-3}\)

\(1.10\cdot 10^{-3}\)

\(1.62\cdot 10^{-3}\)

\(9.11\cdot 10^{-4}\)

\(\mathcal {E}(\phi , \hat{\phi })\)

\(2.67 \cdot 10^{-5}\)

\(6.65\cdot 10^{-6}\)

\(5.93\cdot 10^{-6}\)

\(3.64\cdot 10^{-6}\)

\(3.97\cdot 10^{-6}\)

\(2.52\cdot 10^{-6}\)

\(6.25\cdot 10^{-5}\)

\(1.65\cdot 10^{-4}\)

\(4.87\cdot 10^{-5}\)

\(7.80\cdot 10^{-5}\)

\(4.04\cdot 10^{-5}\)