Table 2 Performance boost of PINN-anneal on the Helmholtz benchmarks using transfer learning
From: Transfer learning for deep neural network-based partial differential equations solving
Subtask | Original | Transfer | Boost |
|---|---|---|---|
case 2 | 1.87e+00 | 4.99e-02 | 97.3% |
case 3 | 2.95e-02 | 1.47e-02 | 50.2% |
case 4 | 3.88e-02 | 2.49e-02 | 35.8% |
case 5 | 4.17e-01 | 6.06e-02 | 85.5% |
case 6 | 7.46e-01 | 2.18e-01 | 70.8% |
case 7 | 1.47e+00 | 9.92e-01 | 32.5% |