Figure 7
From: A machine learning framework for rapid forecasting and history matching in unconventional reservoirs
Parameter prediction from the ANN with parameter loss function (7) before (left) and after (right) transfer learning. Observe that training with synthetic data from the Patzek reduced-order model yields reasonable performance which improves remarkably after transfer learning. The distribution of the true parameters has a mean and standard deviation of 0.07 and 0.23 for \(\bar{\alpha }\) and 0.49 and 0.22 for \(\bar{M}\).
