Regressor validation with scikit-learn ≥ 1.6

[NGC2023]

Hi,
I have created a patch to address the AttributeError with scikit-learn ≥ 1.6 new tag system.

--- a/skopt/optimizer/optimizer.py
+++ b/skopt/optimizer/optimizer.py
@@ -253,8 +253,13 @@ class Optimizer:
             )
 
         # check if regressor
-        if not is_regressor(base_estimator) and base_estimator is not None:
-            raise ValueError("%s has to be a regressor." % base_estimator)
+        if base_estimator is not None and not isinstance(base_estimator, str):
+            try:
+                ok = is_regressor(base_estimator)
+            except Exception:
+                ok = False
+            if not ok:
+                raise ValueError("base_estimator has to be a regressor.")
 
         # treat per second acqusition function specially
         is_multi_regressor = isinstance(base_estimator, MultiOutputRegressor)