Merge pull request #105 from deepskies/issue/DER_loss

modified to be DER_wst with a scaling by this term
deepskies · May 17, 2024 · dae46d4 · dae46d4
2 parents 52a8436 + ec3c92c
commit dae46d4
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 4 deletions.
diff --git a/src/models/models.py b/src/models/models.py
@@ -189,7 +189,7 @@ def loss_der(y, y_pred, coeff):
     gamma, nu, alpha, beta = y[:, 0], y[:, 1], y[:, 2], y[:, 3]
     error = gamma - y_pred
     omega = 2.0 * beta * (1.0 + nu)
-
+    w_st = torch.sqrt(beta * (1 + nu) / (alpha * nu))
     # define aleatoric and epistemic uncert
     u_al = np.sqrt(
         beta.detach().numpy()
@@ -204,7 +204,7 @@ def loss_der(y, y_pred, coeff):
             + (alpha + 0.5) * torch.log(error**2 * nu + omega)
             + torch.lgamma(alpha)
             - torch.lgamma(alpha + 0.5)
-            + coeff * torch.abs(error) * (2.0 * nu + alpha)
+            + (coeff * torch.abs(error / w_st) * (2.0 * nu + alpha))
         ),
         u_al,
         u_ep,

diff --git a/src/scripts/Aleatoric.py b/src/scripts/Aleatoric.py
@@ -290,7 +290,7 @@ def beta_type(value):
             ax.set_title("Deep Evidential Regression")
         elif model[0:2] == "DE":
             ax.set_title("Deep Ensemble (100 models)")
-        ax.set_ylim([0, 14])
+        ax.set_ylim([0, 11])
     plt.legend()
     if config.get_item("analysis", "savefig", "Analysis"):
         plt.savefig(

diff --git a/src/utils/defaults.py b/src/utils/defaults.py
@@ -100,7 +100,7 @@
     },
     "analysis": {
         "noise_level_list": ["low", "medium", "high"],
-        "model_names_list": ["DER", "DE_desiderata_2"],
+        "model_names_list": ["DER_wst", "DE_desiderata_2"],
         # ["DER_desiderata_2", "DE_desiderata_2"]
         "plot": True,
         "savefig": False,