diff --git a/azure-pipelines.yml b/azure-pipelines.yml
index 0263c4d1d..5c4218dec 100644
--- a/azure-pipelines.yml
+++ b/azure-pipelines.yml
@@ -115,10 +115,10 @@ jobs:
         ne(variables['Build.Reason'], 'Schedule')
       )
     matrix:
-      py38_conda_forge_openblas_ubuntu_1804:
+      py39_conda_forge_openblas_ubuntu_1804:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.8'
+        PYTHON_VERSION: '3.9'
         BLAS: 'openblas'
         COVERAGE: 'false'
 
@@ -188,7 +188,7 @@ jobs:
       pylatest_conda_tensorflow:
         DISTRIB: 'conda-latest-tensorflow'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.8'
+        PYTHON_VERSION: '3.9'
         TEST_DOCS: 'true'
         TEST_DOCSTRINGS: 'true'
         CHECK_WARNINGS: 'true'
@@ -214,7 +214,7 @@ jobs:
       pylatest_conda_keras:
         DISTRIB: 'conda-latest-keras'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.8'
+        PYTHON_VERSION: '3.9'
         TEST_DOCS: 'true'
         TEST_DOCSTRINGS: 'true'
         CHECK_WARNINGS: 'true'
@@ -301,7 +301,7 @@ jobs:
       py38_conda_forge_mkl:
         DISTRIB: 'conda'
         CONDA_CHANNEL: 'conda-forge'
-        PYTHON_VERSION: '3.8'
+        PYTHON_VERSION: '3.10'
         CHECK_WARNINGS: 'true'
         PYTHON_ARCH: '64'
         PYTEST_VERSION: '*'
diff --git a/doc/ensemble.rst b/doc/ensemble.rst
index 8efd73947..d8b6751d3 100644
--- a/doc/ensemble.rst
+++ b/doc/ensemble.rst
@@ -33,8 +33,7 @@ data set, this classifier will favor the majority classes::
   >>> from sklearn.ensemble import BaggingClassifier
   >>> from sklearn.tree import DecisionTreeClassifier
   >>> X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0)
-  >>> bc = BaggingClassifier(base_estimator=DecisionTreeClassifier(),
-  ...                        random_state=0)
+  >>> bc = BaggingClassifier(DecisionTreeClassifier(), random_state=0)
   >>> bc.fit(X_train, y_train) #doctest:
   BaggingClassifier(...)
   >>> y_pred = bc.predict(X_test)
@@ -50,7 +49,7 @@ sampling is controlled by the parameter `sampler` or the two parameters
 :class:`~imblearn.under_sampling.RandomUnderSampler`::
 
   >>> from imblearn.ensemble import BalancedBaggingClassifier
-  >>> bbc = BalancedBaggingClassifier(base_estimator=DecisionTreeClassifier(),
+  >>> bbc = BalancedBaggingClassifier(DecisionTreeClassifier(),
   ...                                 sampling_strategy='auto',
   ...                                 replacement=False,
   ...                                 random_state=0)
diff --git a/doc/whats_new/v0.12.rst b/doc/whats_new/v0.12.rst
index 08172f829..1c4325356 100644
--- a/doc/whats_new/v0.12.rst
+++ b/doc/whats_new/v0.12.rst
@@ -23,9 +23,13 @@ Compatibility
 
 - :class:`~imblearn.ensemble.BalancedRandomForestClassifier` now support missing values
   and monotonic constraints if scikit-learn >= 1.4 is installed.
+
 - :class:`~imblearn.pipeline.Pipeline` support metadata routing if scikit-learn >= 1.4
   is installed.
 
+- Compatibility with scikit-learn 1.4.
+  :pr:`1058` by :user:`Guillaume Lemaitre <glemaitre>`.
+
 Deprecations
 ............
 
diff --git a/imblearn/ensemble/_bagging.py b/imblearn/ensemble/_bagging.py
index 04d2217b9..afcf3fd3a 100644
--- a/imblearn/ensemble/_bagging.py
+++ b/imblearn/ensemble/_bagging.py
@@ -5,7 +5,6 @@
 # License: MIT
 
 import copy
-import inspect
 import numbers
 import warnings
 
@@ -15,6 +14,7 @@
 from sklearn.ensemble import BaggingClassifier
 from sklearn.ensemble._bagging import _parallel_decision_function
 from sklearn.ensemble._base import _partition_estimators
+from sklearn.exceptions import NotFittedError
 from sklearn.tree import DecisionTreeClassifier
 from sklearn.utils import parse_version
 from sklearn.utils.validation import check_is_fitted
@@ -121,14 +121,6 @@ class BalancedBaggingClassifier(_ParamsValidationMixin, BaggingClassifier):
 
         .. versionadded:: 0.8
 
-    base_estimator : estimator object, default=None
-        The base estimator to fit on random subsets of the dataset.
-        If None, then the base estimator is a decision tree.
-
-        .. deprecated:: 0.10
-           `base_estimator` was renamed to `estimator` in version 0.10 and
-           will be removed in 0.12.
-
     Attributes
     ----------
     estimator_ : estimator
@@ -136,15 +128,6 @@ class BalancedBaggingClassifier(_ParamsValidationMixin, BaggingClassifier):
 
         .. versionadded:: 0.10
 
-    base_estimator_ : estimator
-        The base estimator from which the ensemble is grown.
-
-        .. deprecated:: 1.2
-           `base_estimator_` is deprecated in `scikit-learn` 1.2 and will be
-           removed in 1.4. Use `estimator_` instead. When the minimum version
-           of `scikit-learn` supported by `imbalanced-learn` will reach 1.4,
-           this attribute will be removed.
-
     n_features_ : int
         The number of features when `fit` is performed.
 
@@ -266,7 +249,7 @@ class BalancedBaggingClassifier(_ParamsValidationMixin, BaggingClassifier):
     """
 
     # make a deepcopy to not modify the original dictionary
-    if sklearn_version >= parse_version("1.3"):
+    if sklearn_version >= parse_version("1.4"):
         _parameter_constraints = copy.deepcopy(BaggingClassifier._parameter_constraints)
     else:
         _parameter_constraints = copy.deepcopy(_bagging_parameter_constraints)
@@ -283,6 +266,9 @@ class BalancedBaggingClassifier(_ParamsValidationMixin, BaggingClassifier):
             "sampler": [HasMethods(["fit_resample"]), None],
         }
     )
+    # TODO: remove when minimum supported version of scikit-learn is 1.4
+    if "base_estimator" in _parameter_constraints:
+        del _parameter_constraints["base_estimator"]
 
     def __init__(
         self,
@@ -301,18 +287,8 @@ def __init__(
         random_state=None,
         verbose=0,
         sampler=None,
-        base_estimator="deprecated",
     ):
-        # TODO: remove when supporting scikit-learn>=1.2
-        bagging_classifier_signature = inspect.signature(super().__init__)
-        estimator_params = {"base_estimator": base_estimator}
-        if "estimator" in bagging_classifier_signature.parameters:
-            estimator_params["estimator"] = estimator
-        else:
-            self.estimator = estimator
-
         super().__init__(
-            **estimator_params,
             n_estimators=n_estimators,
             max_samples=max_samples,
             max_features=max_features,
@@ -324,6 +300,7 @@ def __init__(
             random_state=random_state,
             verbose=verbose,
         )
+        self.estimator = estimator
         self.sampling_strategy = sampling_strategy
         self.replacement = replacement
         self.sampler = sampler
@@ -349,42 +326,17 @@ def _validate_y(self, y):
     def _validate_estimator(self, default=DecisionTreeClassifier()):
         """Check the estimator and the n_estimator attribute, set the
         `estimator_` attribute."""
-        if self.estimator is not None and (
-            self.base_estimator not in [None, "deprecated"]
-        ):
-            raise ValueError(
-                "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-            )
-
         if self.estimator is not None:
-            base_estimator = clone(self.estimator)
-        elif self.base_estimator not in [None, "deprecated"]:
-            warnings.warn(
-                "`base_estimator` was renamed to `estimator` in version 0.10 and "
-                "will be removed in 0.12.",
-                FutureWarning,
-            )
-            base_estimator = clone(self.base_estimator)
+            estimator = clone(self.estimator)
         else:
-            base_estimator = clone(default)
+            estimator = clone(default)
 
         if self.sampler_._sampling_type != "bypass":
             self.sampler_.set_params(sampling_strategy=self._sampling_strategy)
 
-        self._estimator = Pipeline(
-            [("sampler", self.sampler_), ("classifier", base_estimator)]
+        self.estimator_ = Pipeline(
+            [("sampler", self.sampler_), ("classifier", estimator)]
         )
-        try:
-            # scikit-learn < 1.2
-            self.base_estimator_ = self._estimator
-        except AttributeError:
-            pass
-
-    # TODO: remove when supporting scikit-learn>=1.4
-    @property
-    def estimator_(self):
-        """Estimator used to grow the ensemble."""
-        return self._estimator
 
     # TODO: remove when supporting scikit-learn>=1.2
     @property
@@ -483,6 +435,22 @@ def decision_function(self, X):
 
         return decisions
 
+    @property
+    def base_estimator_(self):
+        """Attribute for older sklearn version compatibility."""
+        error = AttributeError(
+            f"{self.__class__.__name__} object has no attribute 'base_estimator_'."
+        )
+        if sklearn_version < parse_version("1.2"):
+            # The base class require to have the attribute defined. For scikit-learn
+            # > 1.2, we are going to raise an error.
+            try:
+                check_is_fitted(self)
+                return self.estimator_
+            except NotFittedError:
+                raise error
+        raise error
+
     def _more_tags(self):
         tags = super()._more_tags()
         tags_key = "_xfail_checks"
diff --git a/imblearn/ensemble/_easy_ensemble.py b/imblearn/ensemble/_easy_ensemble.py
index e2f5575cb..db3c6cbb7 100644
--- a/imblearn/ensemble/_easy_ensemble.py
+++ b/imblearn/ensemble/_easy_ensemble.py
@@ -5,7 +5,6 @@
 # License: MIT
 
 import copy
-import inspect
 import numbers
 import warnings
 
@@ -15,6 +14,7 @@
 from sklearn.ensemble import AdaBoostClassifier, BaggingClassifier
 from sklearn.ensemble._bagging import _parallel_decision_function
 from sklearn.ensemble._base import _partition_estimators
+from sklearn.exceptions import NotFittedError
 from sklearn.utils import parse_version
 from sklearn.utils._tags import _safe_tags
 from sklearn.utils.validation import check_is_fitted
@@ -85,14 +85,6 @@ class EasyEnsembleClassifier(_ParamsValidationMixin, BaggingClassifier):
     verbose : int, default=0
         Controls the verbosity of the building process.
 
-    base_estimator : estimator object, default=AdaBoostClassifier()
-        The base AdaBoost classifier used in the inner ensemble. Note that you
-        can set the number of inner learner by passing your own instance.
-
-        .. deprecated:: 0.10
-           `base_estimator` was renamed to `estimator` in version 0.10 and will
-           be removed in 0.12.
-
     Attributes
     ----------
     estimator_ : estimator
@@ -100,15 +92,6 @@ class EasyEnsembleClassifier(_ParamsValidationMixin, BaggingClassifier):
 
         .. versionadded:: 0.10
 
-    base_estimator_ : estimator
-        The base estimator from which the ensemble is grown.
-
-        .. deprecated:: 1.2
-           `base_estimator_` is deprecated in `scikit-learn` 1.2 and will be
-           removed in 1.4. Use `estimator_` instead. When the minimum version
-           of `scikit-learn` supported by `imbalanced-learn` will reach 1.4,
-           this attribute will be removed.
-
     estimators_ : list of estimators
         The collection of fitted base estimators.
 
@@ -191,7 +174,7 @@ class EasyEnsembleClassifier(_ParamsValidationMixin, BaggingClassifier):
     """
 
     # make a deepcopy to not modify the original dictionary
-    if sklearn_version >= parse_version("1.3"):
+    if sklearn_version >= parse_version("1.4"):
         _parameter_constraints = copy.deepcopy(BaggingClassifier._parameter_constraints)
     else:
         _parameter_constraints = copy.deepcopy(_bagging_parameter_constraints)
@@ -217,6 +200,9 @@ class EasyEnsembleClassifier(_ParamsValidationMixin, BaggingClassifier):
             "replacement": ["boolean"],
         }
     )
+    # TODO: remove when minimum supported version of scikit-learn is 1.4
+    if "base_estimator" in _parameter_constraints:
+        del _parameter_constraints["base_estimator"]
 
     def __init__(
         self,
@@ -229,18 +215,8 @@ def __init__(
         n_jobs=None,
         random_state=None,
         verbose=0,
-        base_estimator="deprecated",
     ):
-        # TODO: remove when supporting scikit-learn>=1.2
-        bagging_classifier_signature = inspect.signature(super().__init__)
-        estimator_params = {"base_estimator": base_estimator}
-        if "estimator" in bagging_classifier_signature.parameters:
-            estimator_params["estimator"] = estimator
-        else:
-            self.estimator = estimator
-
         super().__init__(
-            **estimator_params,
             n_estimators=n_estimators,
             max_samples=1.0,
             max_features=1.0,
@@ -252,6 +228,7 @@ def __init__(
             random_state=random_state,
             verbose=verbose,
         )
+        self.estimator = estimator
         self.sampling_strategy = sampling_strategy
         self.replacement = replacement
 
@@ -270,46 +247,19 @@ def _validate_y(self, y):
             self._sampling_strategy = self.sampling_strategy
         return y_encoded
 
-    def _validate_estimator(self, default=AdaBoostClassifier()):
+    def _validate_estimator(self, default=AdaBoostClassifier(algorithm="SAMME")):
         """Check the estimator and the n_estimator attribute, set the
         `estimator_` attribute."""
-        if self.estimator is not None and (
-            self.base_estimator not in [None, "deprecated"]
-        ):
-            raise ValueError(
-                "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-            )
-
         if self.estimator is not None:
-            base_estimator = clone(self.estimator)
-        elif self.base_estimator not in [None, "deprecated"]:
-            warnings.warn(
-                "`base_estimator` was renamed to `estimator` in version 0.10 and "
-                "will be removed in 0.12.",
-                FutureWarning,
-            )
-            base_estimator = clone(self.base_estimator)
+            estimator = clone(self.estimator)
         else:
-            base_estimator = clone(default)
+            estimator = clone(default)
 
         sampler = RandomUnderSampler(
             sampling_strategy=self._sampling_strategy,
             replacement=self.replacement,
         )
-        self._estimator = Pipeline(
-            [("sampler", sampler), ("classifier", base_estimator)]
-        )
-        try:
-            self.base_estimator_ = self._estimator
-        except AttributeError:
-            # scikit-learn < 1.2
-            pass
-
-    # TODO: remove when supporting scikit-learn>=1.4
-    @property
-    def estimator_(self):
-        """Estimator used to grow the ensemble."""
-        return self._estimator
+        self.estimator_ = Pipeline([("sampler", sampler), ("classifier", estimator)])
 
     # TODO: remove when supporting scikit-learn>=1.2
     @property
@@ -399,9 +349,25 @@ def decision_function(self, X):
 
         return decisions
 
+    @property
+    def base_estimator_(self):
+        """Attribute for older sklearn version compatibility."""
+        error = AttributeError(
+            f"{self.__class__.__name__} object has no attribute 'base_estimator_'."
+        )
+        if sklearn_version < parse_version("1.2"):
+            # The base class require to have the attribute defined. For scikit-learn
+            # > 1.2, we are going to raise an error.
+            try:
+                check_is_fitted(self)
+                return self.estimator_
+            except NotFittedError:
+                raise error
+        raise error
+
     def _more_tags(self):
         if self.estimator is None:
-            estimator = AdaBoostClassifier()
+            estimator = AdaBoostClassifier(algorithm="SAMME")
         else:
             estimator = self.estimator
         return {"allow_nan": _safe_tags(estimator, "allow_nan")}
diff --git a/imblearn/ensemble/_forest.py b/imblearn/ensemble/_forest.py
index b8ef60c6e..a7c8f9beb 100644
--- a/imblearn/ensemble/_forest.py
+++ b/imblearn/ensemble/_forest.py
@@ -334,16 +334,6 @@ class BalancedRandomForestClassifier(_ParamsValidationMixin, RandomForestClassif
 
         .. versionadded:: 0.10
 
-    base_estimator_ : :class:`~sklearn.tree.DecisionTreeClassifier` instance
-        The child estimator template used to create the collection of fitted
-        sub-estimators.
-
-        .. deprecated:: 1.2
-           `base_estimator_` is deprecated in `scikit-learn` 1.2 and will be
-           removed in 1.4. Use `estimator_` instead. When the minimum version
-           of `scikit-learn` supported by `imbalanced-learn` will reach 1.4,
-           this attribute will be removed.
-
     estimators_ : list of :class:`~sklearn.tree.DecisionTreeClassifier`
         The collection of fitted sub-estimators.
 
@@ -529,15 +519,9 @@ def _validate_estimator(self, default=DecisionTreeClassifier()):
             base_estimator = self.base_estimator
 
         if base_estimator is not None:
-            self._estimator = clone(base_estimator)
+            self.estimator_ = clone(base_estimator)
         else:
-            self._estimator = clone(default)
-
-        try:
-            # scikit-learn < 1.2
-            self.base_estimator_ = self._estimator
-        except AttributeError:
-            pass
+            self.estimator_ = clone(default)
 
         self.base_sampler_ = RandomUnderSampler(
             sampling_strategy=self._sampling_strategy,
@@ -549,7 +533,7 @@ def _make_sampler_estimator(self, random_state=None):
         Warning: This method should be used to properly instantiate new
         sub-estimators.
         """
-        estimator = clone(self._estimator)
+        estimator = clone(self.estimator_)
         estimator.set_params(**{p: getattr(self, p) for p in self.estimator_params})
         sampler = clone(self.base_sampler_)
 
@@ -910,12 +894,6 @@ def _compute_oob_predictions(self, X, y):
 
         return oob_pred
 
-    # TODO: remove when supporting scikit-learn>=1.4
-    @property
-    def estimator_(self):
-        """Estimator used to grow the ensemble."""
-        return self._estimator
-
     # TODO: remove when supporting scikit-learn>=1.2
     @property
     def n_features_(self):
diff --git a/imblearn/ensemble/_weight_boosting.py b/imblearn/ensemble/_weight_boosting.py
index 7f7965d5e..539b7824f 100644
--- a/imblearn/ensemble/_weight_boosting.py
+++ b/imblearn/ensemble/_weight_boosting.py
@@ -1,7 +1,5 @@
 import copy
-import inspect
 import numbers
-import warnings
 from copy import deepcopy
 
 import numpy as np
@@ -66,6 +64,10 @@ class RUSBoostClassifier(_ParamsValidationMixin, AdaBoostClassifier):
         The SAMME.R algorithm typically converges faster than SAMME,
         achieving a lower test error with fewer boosting iterations.
 
+        .. deprecated:: 0.12
+            `"SAMME.R"` is deprecated and will be removed in version 0.14.
+            '"SAMME"' will become the default.
+
     {sampling_strategy}
 
     replacement : bool, default=False
@@ -73,16 +75,6 @@ class RUSBoostClassifier(_ParamsValidationMixin, AdaBoostClassifier):
 
     {random_state}
 
-    base_estimator : estimator object, default=None
-        The base estimator from which the boosted ensemble is built.
-        Support for sample weighting is required, as well as proper
-        ``classes_`` and ``n_classes_`` attributes. If ``None``, then
-        the base estimator is ``DecisionTreeClassifier(max_depth=1)``.
-
-        .. deprecated:: 0.10
-           `base_estimator` is deprecated in version 0.10 and will be removed
-           in 0.12. Use `estimator` instead.
-
     Attributes
     ----------
     estimator_ : estimator
@@ -90,15 +82,6 @@ class RUSBoostClassifier(_ParamsValidationMixin, AdaBoostClassifier):
 
         .. versionadded:: 0.10
 
-    base_estimator_ : estimator
-        The base estimator from which the ensemble is grown.
-
-        .. deprecated:: 1.2
-           `base_estimator_` is deprecated in `scikit-learn` 1.2 and will be
-           removed in 1.4. Use `estimator_` instead. When the minimum version
-           of `scikit-learn` supported by `imbalanced-learn` will reach 1.4,
-           this attribute will be removed.
-
     estimators_ : list of classifiers
         The collection of fitted sub-estimators.
 
@@ -172,7 +155,7 @@ class RUSBoostClassifier(_ParamsValidationMixin, AdaBoostClassifier):
     """
 
     # make a deepcopy to not modify the original dictionary
-    if sklearn_version >= parse_version("1.3"):
+    if sklearn_version >= parse_version("1.4"):
         _parameter_constraints = copy.deepcopy(
             AdaBoostClassifier._parameter_constraints
         )
@@ -192,6 +175,9 @@ class RUSBoostClassifier(_ParamsValidationMixin, AdaBoostClassifier):
             "replacement": ["boolean"],
         }
     )
+    # TODO: remove when minimum supported version of scikit-learn is 1.4
+    if "base_estimator" in _parameter_constraints:
+        del _parameter_constraints["base_estimator"]
 
     def __init__(
         self,
@@ -203,23 +189,14 @@ def __init__(
         sampling_strategy="auto",
         replacement=False,
         random_state=None,
-        base_estimator="deprecated",
     ):
-        # TODO: remove when supporting scikit-learn>=1.2
-        bagging_classifier_signature = inspect.signature(super().__init__)
-        estimator_params = {"base_estimator": base_estimator}
-        if "estimator" in bagging_classifier_signature.parameters:
-            estimator_params["estimator"] = estimator
-        else:
-            self.estimator = estimator
-
         super().__init__(
-            **estimator_params,
             n_estimators=n_estimators,
             learning_rate=learning_rate,
             algorithm=algorithm,
             random_state=random_state,
         )
+        self.estimator = estimator
         self.sampling_strategy = sampling_strategy
         self.replacement = replacement
 
@@ -257,36 +234,15 @@ def _validate_estimator(self):
 
         Sets the `estimator_` attributes.
         """
-        if self.estimator is not None and (
-            self.base_estimator not in [None, "deprecated"]
-        ):
-            raise ValueError(
-                "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-            )
-
         default = DecisionTreeClassifier(max_depth=1)
         if self.estimator is not None:
-            base_estimator = clone(self.estimator)
-        elif self.base_estimator not in [None, "deprecated"]:
-            warnings.warn(
-                "`base_estimator` was renamed to `estimator` in version 0.10 and "
-                "will be removed in 0.12.",
-                FutureWarning,
-            )
-            base_estimator = clone(self.base_estimator)
+            self.estimator_ = clone(self.estimator)
         else:
-            base_estimator = clone(default)
-
-        self._estimator = base_estimator
-        try:
-            # scikit-learn < 1.2
-            self.base_estimator_ = self._estimator
-        except AttributeError:
-            pass
+            self.estimator_ = clone(default)
 
         #  SAMME-R requires predict_proba-enabled estimators
         if self.algorithm == "SAMME.R":
-            if not hasattr(self._estimator, "predict_proba"):
+            if not hasattr(self.estimator_, "predict_proba"):
                 raise TypeError(
                     "AdaBoostClassifier with algorithm='SAMME.R' requires "
                     "that the weak learner supports the calculation of class "
@@ -294,9 +250,9 @@ def _validate_estimator(self):
                     "Please change the base estimator or set "
                     "algorithm='SAMME' instead."
                 )
-        if not has_fit_parameter(self._estimator, "sample_weight"):
+        if not has_fit_parameter(self.estimator_, "sample_weight"):
             raise ValueError(
-                f"{self._estimator.__class__.__name__} doesn't support sample_weight."
+                f"{self.estimator_.__class__.__name__} doesn't support sample_weight."
             )
 
         self.base_sampler_ = RandomUnderSampler(
@@ -309,7 +265,7 @@ def _make_sampler_estimator(self, append=True, random_state=None):
         Warning: This method should be used to properly instantiate new
         sub-estimators.
         """
-        estimator = clone(self._estimator)
+        estimator = clone(self.estimator_)
         estimator.set_params(**{p: getattr(self, p) for p in self.estimator_params})
         sampler = clone(self.base_sampler_)
 
@@ -437,9 +393,3 @@ def _boost_discrete(self, iboost, X, y, sample_weight, random_state):
             sample_weight *= np.exp(estimator_weight * incorrect * (sample_weight > 0))
 
         return sample_weight, estimator_weight, estimator_error
-
-    # TODO: remove when supporting scikit-learn>=1.4
-    @property
-    def estimator_(self):
-        """Estimator used to grow the ensemble."""
-        return self._estimator
diff --git a/imblearn/ensemble/tests/test_bagging.py b/imblearn/ensemble/tests/test_bagging.py
index 01532add8..5705de553 100644
--- a/imblearn/ensemble/tests/test_bagging.py
+++ b/imblearn/ensemble/tests/test_bagging.py
@@ -592,25 +592,3 @@ def test_balanced_bagging_classifier_n_features():
     estimator = BalancedBaggingClassifier().fit(X, y)
     with pytest.warns(FutureWarning, match="`n_features_` was deprecated"):
         estimator.n_features_
-
-
-@pytest.mark.skipif(
-    sklearn_version < parse_version("1.2"), reason="requires scikit-learn>=1.2"
-)
-def test_balanced_bagging_classifier_base_estimator():
-    """Check that we raise a FutureWarning when accessing `base_estimator_`."""
-    X, y = load_iris(return_X_y=True)
-    estimator = BalancedBaggingClassifier().fit(X, y)
-    with pytest.warns(FutureWarning, match="`base_estimator_` was deprecated"):
-        estimator.base_estimator_
-
-
-def test_balanced_bagging_classifier_set_both_estimator_and_base_estimator():
-    """Check that we raise a ValueError when setting both `estimator` and
-    `base_estimator`."""
-    X, y = load_iris(return_X_y=True)
-    err_msg = "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-    with pytest.raises(ValueError, match=err_msg):
-        BalancedBaggingClassifier(
-            estimator=KNeighborsClassifier(), base_estimator=KNeighborsClassifier()
-        ).fit(X, y)
diff --git a/imblearn/ensemble/tests/test_easy_ensemble.py b/imblearn/ensemble/tests/test_easy_ensemble.py
index a8574e4d6..7dc04414a 100644
--- a/imblearn/ensemble/tests/test_easy_ensemble.py
+++ b/imblearn/ensemble/tests/test_easy_ensemble.py
@@ -43,7 +43,10 @@
 @pytest.mark.parametrize("n_estimators", [10, 20])
 @pytest.mark.parametrize(
     "estimator",
-    [AdaBoostClassifier(n_estimators=5), AdaBoostClassifier(n_estimators=10)],
+    [
+        AdaBoostClassifier(algorithm="SAMME", n_estimators=5),
+        AdaBoostClassifier(algorithm="SAMME", n_estimators=10),
+    ],
 )
 def test_easy_ensemble_classifier(n_estimators, estimator):
     # Check classification for various parameter settings.
@@ -89,7 +92,7 @@ def test_estimator():
     assert isinstance(ensemble.estimator_.steps[-1][1], AdaBoostClassifier)
 
     ensemble = EasyEnsembleClassifier(
-        2, AdaBoostClassifier(), n_jobs=-1, random_state=0
+        2, AdaBoostClassifier(algorithm="SAMME"), n_jobs=-1, random_state=0
     ).fit(X_train, y_train)
 
     assert isinstance(ensemble.estimator_.steps[-1][1], AdaBoostClassifier)
@@ -104,7 +107,9 @@ def test_bagging_with_pipeline():
     )
     estimator = EasyEnsembleClassifier(
         n_estimators=2,
-        estimator=make_pipeline(SelectKBest(k=1), AdaBoostClassifier()),
+        estimator=make_pipeline(
+            SelectKBest(k=1), AdaBoostClassifier(algorithm="SAMME")
+        ),
     )
     estimator.fit(X, y).predict(X)
 
@@ -195,7 +200,8 @@ def test_easy_ensemble_classifier_single_estimator():
 
     clf1 = EasyEnsembleClassifier(n_estimators=1, random_state=0).fit(X_train, y_train)
     clf2 = make_pipeline(
-        RandomUnderSampler(random_state=0), AdaBoostClassifier(random_state=0)
+        RandomUnderSampler(random_state=0),
+        AdaBoostClassifier(algorithm="SAMME", random_state=0),
     ).fit(X_train, y_train)
 
     assert_array_equal(clf1.predict(X_test), clf2.predict(X_test))
@@ -214,7 +220,7 @@ def test_easy_ensemble_classifier_grid_search():
         "estimator__n_estimators": [3, 4],
     }
     grid_search = GridSearchCV(
-        EasyEnsembleClassifier(estimator=AdaBoostClassifier()),
+        EasyEnsembleClassifier(estimator=AdaBoostClassifier(algorithm="SAMME")),
         parameters,
         cv=5,
     )
@@ -227,25 +233,3 @@ def test_easy_ensemble_classifier_n_features():
     estimator = EasyEnsembleClassifier().fit(X, y)
     with pytest.warns(FutureWarning, match="`n_features_` was deprecated"):
         estimator.n_features_
-
-
-@pytest.mark.skipif(
-    sklearn_version < parse_version("1.2"), reason="warns for scikit-learn>=1.2"
-)
-def test_easy_ensemble_classifier_base_estimator():
-    """Check that we raise a FutureWarning when accessing `base_estimator_`."""
-    X, y = load_iris(return_X_y=True)
-    estimator = EasyEnsembleClassifier().fit(X, y)
-    with pytest.warns(FutureWarning, match="`base_estimator_` was deprecated"):
-        estimator.base_estimator_
-
-
-def test_easy_ensemble_classifier_set_both_estimator_and_base_estimator():
-    """Check that we raise a ValueError when setting both `estimator` and
-    `base_estimator`."""
-    X, y = load_iris(return_X_y=True)
-    err_msg = "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-    with pytest.raises(ValueError, match=err_msg):
-        EasyEnsembleClassifier(
-            estimator=AdaBoostClassifier(), base_estimator=AdaBoostClassifier()
-        ).fit(X, y)
diff --git a/imblearn/ensemble/tests/test_forest.py b/imblearn/ensemble/tests/test_forest.py
index 9bd73de65..3719568e5 100644
--- a/imblearn/ensemble/tests/test_forest.py
+++ b/imblearn/ensemble/tests/test_forest.py
@@ -229,19 +229,6 @@ def test_balanced_bagging_classifier_n_features():
         estimator.n_features_
 
 
-@pytest.mark.skipif(
-    sklearn_version < parse_version("1.2"), reason="requires scikit-learn>=1.2"
-)
-def test_balanced_random_forest_classifier_base_estimator():
-    """Check that we raise a FutureWarning when accessing `base_estimator_`."""
-    X, y = load_iris(return_X_y=True)
-    estimator = BalancedRandomForestClassifier(
-        sampling_strategy="all", replacement=True, bootstrap=False
-    ).fit(X, y)
-    with pytest.warns(FutureWarning, match="`base_estimator_` was deprecated"):
-        estimator.base_estimator_
-
-
 # TODO: remove in 0.13
 def test_balanced_random_forest_change_behaviour(imbalanced_dataset):
     """Check that we raise a change of behaviour for the parameters `sampling_strategy`
diff --git a/imblearn/ensemble/tests/test_weight_boosting.py b/imblearn/ensemble/tests/test_weight_boosting.py
index a36395e55..ad3dbca04 100644
--- a/imblearn/ensemble/tests/test_weight_boosting.py
+++ b/imblearn/ensemble/tests/test_weight_boosting.py
@@ -1,9 +1,8 @@
 import numpy as np
 import pytest
 import sklearn
-from sklearn.datasets import load_iris, make_classification
+from sklearn.datasets import make_classification
 from sklearn.model_selection import train_test_split
-from sklearn.tree import DecisionTreeClassifier
 from sklearn.utils._testing import assert_array_equal
 from sklearn.utils.fixes import parse_version
 
@@ -29,6 +28,7 @@ def imbalanced_dataset():
 
 
 @pytest.mark.parametrize("algorithm", ["SAMME", "SAMME.R"])
+@pytest.mark.filterwarnings("ignore:The SAMME.R algorithm (the default) is")
 def test_rusboost(imbalanced_dataset, algorithm):
     X, y = imbalanced_dataset
     X_train, X_test, y_train, y_test = train_test_split(
@@ -74,6 +74,7 @@ def test_rusboost(imbalanced_dataset, algorithm):
 
 
 @pytest.mark.parametrize("algorithm", ["SAMME", "SAMME.R"])
+@pytest.mark.filterwarnings("ignore:The SAMME.R algorithm (the default) is")
 def test_rusboost_sample_weight(imbalanced_dataset, algorithm):
     X, y = imbalanced_dataset
     sample_weight = np.ones_like(y)
@@ -91,25 +92,3 @@ def test_rusboost_sample_weight(imbalanced_dataset, algorithm):
 
     with pytest.raises(AssertionError):
         assert_array_equal(y_pred_no_sample_weight, y_pred_sample_weight)
-
-
-@pytest.mark.skipif(
-    sklearn_version < parse_version("1.2"), reason="requires scikit-learn>=1.2"
-)
-def test_rus_boost_classifier_base_estimator():
-    """Check that we raise a FutureWarning when accessing `base_estimator_`."""
-    X, y = load_iris(return_X_y=True)
-    estimator = RUSBoostClassifier().fit(X, y)
-    with pytest.warns(FutureWarning, match="`base_estimator_` was deprecated"):
-        estimator.base_estimator_
-
-
-def test_rus_boost_classifier_set_both_estimator_and_base_estimator():
-    """Check that we raise a ValueError when setting both `estimator` and
-    `base_estimator`."""
-    X, y = load_iris(return_X_y=True)
-    err_msg = "Both `estimator` and `base_estimator` were set. Only set `estimator`."
-    with pytest.raises(ValueError, match=err_msg):
-        RUSBoostClassifier(
-            estimator=DecisionTreeClassifier(), base_estimator=DecisionTreeClassifier()
-        ).fit(X, y)
diff --git a/imblearn/over_sampling/_smote/tests/test_smoten.py b/imblearn/over_sampling/_smote/tests/test_smoten.py
index db9c14f99..2e30e3f19 100644
--- a/imblearn/over_sampling/_smote/tests/test_smoten.py
+++ b/imblearn/over_sampling/_smote/tests/test_smoten.py
@@ -66,7 +66,7 @@ def test_smoten_sparse_input(data, sparse_format):
     https://github.com/scikit-learn-contrib/imbalanced-learn/issues/971
     """
     X, y = data
-    X = OneHotEncoder().fit_transform(X)
+    X = OneHotEncoder().fit_transform(X).toarray()
     X = _convert_container(X, sparse_format)
 
     with pytest.warns(DataConversionWarning, match="is not really efficient"):
diff --git a/imblearn/tests/test_base.py b/imblearn/tests/test_base.py
index 6c16541d4..dc5b705f0 100644
--- a/imblearn/tests/test_base.py
+++ b/imblearn/tests/test_base.py
@@ -26,9 +26,10 @@
 def test_function_sampler_reject_sparse():
     X_sparse = sparse.csr_matrix(X)
     sampler = FunctionSampler(accept_sparse=False)
+    err_msg = "dense data is required"
     with pytest.raises(
         TypeError,
-        match="A sparse matrix was passed, but dense data is required",
+        match=err_msg,
     ):
         sampler.fit_resample(X_sparse, y)
 
diff --git a/imblearn/tests/test_docstring_parameters.py b/imblearn/tests/test_docstring_parameters.py
index 388de5f51..b595d77d7 100644
--- a/imblearn/tests/test_docstring_parameters.py
+++ b/imblearn/tests/test_docstring_parameters.py
@@ -212,7 +212,11 @@ def test_fit_docstring_attributes(name, Estimator):
     else:
         est.fit(X, y)
 
-    skipped_attributes = set([])
+    skipped_attributes = set(
+        [
+            "base_estimator_",  # this attribute exist with old version of sklearn
+        ]
+    )
 
     for attr in attributes:
         if attr.name in skipped_attributes:
diff --git a/imblearn/utils/_param_validation.py b/imblearn/utils/_param_validation.py
index 2bd5e8014..3ccabf2eb 100644
--- a/imblearn/utils/_param_validation.py
+++ b/imblearn/utils/_param_validation.py
@@ -6,7 +6,6 @@
 import math
 import operator
 import re
-import warnings
 from abc import ABC, abstractmethod
 from collections.abc import Iterable
 from inspect import signature
@@ -22,16 +21,14 @@
 
 sklearn_version = parse_version(sklearn.__version__)
 
-if sklearn_version < parse_version("1.3"):
-    # TODO: remove `if True` when we have clear support for:
-    # - ignoring `*args` and `**kwargs` in the signature
+if sklearn_version < parse_version("1.4"):
 
     class InvalidParameterError(ValueError, TypeError):
-        """Custom exception to be raised when the parameter of a
-        class/method/function does not have a valid type or value.
+        """Custom exception to be raised when the parameter of a class/method/function
+        does not have a valid type or value.
         """
 
-    # Inherits from ValueError and TypeError to keep backward compatibility.
+        # Inherits from ValueError and TypeError to keep backward compatibility.
 
     def validate_parameter_constraints(parameter_constraints, params, caller_name):
         """Validate types and values of given parameters.
@@ -56,14 +53,15 @@ def validate_parameter_constraints(parameter_constraints, params, caller_name):
             - the string "boolean"
             - the string "verbose"
             - the string "cv_object"
+            - the string "nan"
             - a MissingValues object representing markers for missing values
             - a HasMethods object, representing method(s) an object must have
             - a Hidden object, representing a constraint not meant to be exposed to the
               user
 
         params : dict
-            A dictionary `param_name: param_value`. The parameters to validate
-            against the constraints.
+            A dictionary `param_name: param_value`. The parameters to validate against
+            the constraints.
 
         caller_name : str
             The name of the estimator or function or method that called this function.
@@ -148,6 +146,8 @@ def make_constraint(constraint):
             constraint = make_constraint(constraint.constraint)
             constraint.hidden = True
             return constraint
+        if isinstance(constraint, str) and constraint == "nan":
+            return _NanConstraint()
         raise ValueError(f"Unknown constraint type: {constraint}")
 
     def validate_params(parameter_constraints, *, prefer_skip_nested_validation):
@@ -156,12 +156,12 @@ def validate_params(parameter_constraints, *, prefer_skip_nested_validation):
         Parameters
         ----------
         parameter_constraints : dict
-            A dictionary `param_name: list of constraints`. See the docstring
-            of `validate_parameter_constraints` for a description of the
-            accepted constraints.
+            A dictionary `param_name: list of constraints`. See the docstring of
+            `validate_parameter_constraints` for a description of the accepted
+            constraints.
 
-            Note that the *args and **kwargs parameters are not validated and
-            must not be present in the parameter_constraints dictionary.
+            Note that the *args and **kwargs parameters are not validated and must not
+            be present in the parameter_constraints dictionary.
 
         prefer_skip_nested_validation : bool
             If True, the validation of parameters of inner estimators or functions
@@ -223,11 +223,10 @@ def wrapper(*args, **kwargs):
                     ):
                         return func(*args, **kwargs)
                 except InvalidParameterError as e:
-                    # When the function is just a wrapper around an estimator,
-                    # we allow the function to delegate validation to the
-                    # estimator, but we replace the name of the estimator by
-                    # the name of the function in the error message to avoid
-                    # confusion.
+                    # When the function is just a wrapper around an estimator, we allow
+                    # the function to delegate validation to the estimator, but we
+                    # replace the name of the estimator by the name of the function in
+                    # the error message to avoid confusion.
                     msg = re.sub(
                         r"parameter of \w+ must be",
                         f"parameter of {func.__qualname__} must be",
@@ -318,7 +317,11 @@ class _NanConstraint(_Constraint):
         """Constraint representing the indicator `np.nan`."""
 
         def is_satisfied_by(self, val):
-            return isinstance(val, Real) and math.isnan(val)
+            return (
+                not isinstance(val, Integral)
+                and isinstance(val, Real)
+                and math.isnan(val)
+            )
 
         def __str__(self):
             return "numpy.nan"
@@ -484,7 +487,7 @@ def _check_params(self):
                 )
 
         def __contains__(self, val):
-            if np.isnan(val):
+            if not isinstance(val, Integral) and np.isnan(val):
                 return False
 
             left_cmp = operator.lt if self.closed in ("left", "both") else operator.le
@@ -586,20 +589,9 @@ def __init__(self):
             self._constraints = [
                 _InstancesOf(bool),
                 _InstancesOf(np.bool_),
-                _InstancesOf(Integral),
             ]
 
         def is_satisfied_by(self, val):
-            # TODO(1.4) remove support for Integral.
-            if isinstance(val, Integral) and not isinstance(val, bool):
-                warnings.warn(
-                    (
-                        "Passing an int for a boolean parameter is deprecated in "
-                        " version 1.2 and won't be supported anymore in version 1.4."
-                    ),
-                    FutureWarning,
-                )
-
             return any(c.is_satisfied_by(val) for c in self._constraints)
 
         def __str__(self):
@@ -680,8 +672,8 @@ def __str__(self):
     class HasMethods(_Constraint):
         """Constraint representing objects that expose specific methods.
 
-        It is useful for parameters following a protocol and where we don't
-        want to impose an affiliation to a specific module or class.
+        It is useful for parameters following a protocol and where we don't want to
+        impose an affiliation to a specific module or class.
 
         Parameters
         ----------
@@ -931,6 +923,7 @@ def generate_valid_param(constraint):
         _CVObjects,
         _InstancesOf,
         _IterablesNotString,
+        _NanConstraint,
         _NoneConstraint,
         _PandasNAConstraint,
         _RandomStates,
diff --git a/imblearn/utils/tests/test_estimator_checks.py b/imblearn/utils/tests/test_estimator_checks.py
index dbc337dd1..ca704f222 100644
--- a/imblearn/utils/tests/test_estimator_checks.py
+++ b/imblearn/utils/tests/test_estimator_checks.py
@@ -101,7 +101,7 @@ def test_check_samplers_nan():
     "BaseBadSampler": (AssertionError, "ValueError not raised by fit"),
     "SamplerSingleClass": (AssertionError, "Sampler can't balance when only"),
     "NotFittedSampler": (AssertionError, "No fitted attribute"),
-    "NoAcceptingSparseSampler": (TypeError, "A sparse matrix was passed"),
+    "NoAcceptingSparseSampler": (TypeError, "dense data is required"),
     "NotPreservingDtypeSampler": (AssertionError, "X dtype is not preserved"),
 }
 
diff --git a/imblearn/utils/tests/test_param_validation.py b/imblearn/utils/tests/test_param_validation.py
index 3a7ab65a3..38af6642d 100644
--- a/imblearn/utils/tests/test_param_validation.py
+++ b/imblearn/utils/tests/test_param_validation.py
@@ -27,6 +27,7 @@
     _CVObjects,
     _InstancesOf,
     _IterablesNotString,
+    _NanConstraint,
     _NoneConstraint,
     _PandasNAConstraint,
     _RandomStates,
@@ -79,16 +80,41 @@ def fit(self, X=None, y=None):
 def test_interval_range(interval_type):
     """Check the range of values depending on closed."""
     interval = Interval(interval_type, -2, 2, closed="left")
-    assert -2 in interval and 2 not in interval
+    assert -2 in interval
+    assert 2 not in interval
 
     interval = Interval(interval_type, -2, 2, closed="right")
-    assert -2 not in interval and 2 in interval
+    assert -2 not in interval
+    assert 2 in interval
 
     interval = Interval(interval_type, -2, 2, closed="both")
-    assert -2 in interval and 2 in interval
+    assert -2 in interval
+    assert 2 in interval
 
     interval = Interval(interval_type, -2, 2, closed="neither")
-    assert -2 not in interval and 2 not in interval
+    assert -2 not in interval
+    assert 2 not in interval
+
+
+@pytest.mark.parametrize("interval_type", [Integral, Real])
+def test_interval_large_integers(interval_type):
+    """Check that Interval constraint work with large integers.
+
+    non-regression test for #26648.
+    """
+    interval = Interval(interval_type, 0, 2, closed="neither")
+    assert 2**65 not in interval
+    assert 2**128 not in interval
+    assert float(2**65) not in interval
+    assert float(2**128) not in interval
+
+    interval = Interval(interval_type, 0, 2**128, closed="neither")
+    assert 2**65 in interval
+    assert 2**128 not in interval
+    assert float(2**65) in interval
+    assert float(2**128) not in interval
+
+    assert 2**1024 not in interval
 
 
 def test_interval_inf_in_bounds():
@@ -392,8 +418,10 @@ def test_generate_valid_param(constraint):
         (Real, 0.5),
         ("boolean", False),
         ("verbose", 1),
+        ("nan", np.nan),
         (MissingValues(), -1),
         (MissingValues(), -1.0),
+        (MissingValues(), 2**1028),
         (MissingValues(), None),
         (MissingValues(), float("nan")),
         (MissingValues(), np.nan),
@@ -425,10 +453,11 @@ def test_is_satisfied_by(constraint_declaration, value):
         (MissingValues(numeric_only=True), MissingValues),
         (HasMethods("fit"), HasMethods),
         ("cv_object", _CVObjects),
+        ("nan", _NanConstraint),
     ],
 )
 def test_make_constraint(constraint_declaration, expected_constraint_class):
-    """Check that make_constraint dispaches to the appropriate constraint class"""
+    """Check that make_constraint dispatches to the appropriate constraint class"""
     constraint = make_constraint(constraint_declaration)
     assert constraint.__class__ is expected_constraint_class
 
@@ -603,12 +632,6 @@ def f(param):
     f(True)
     f(np.bool_(False))
 
-    # an int is also valid but deprecated
-    with pytest.warns(
-        FutureWarning, match="Passing an int for a boolean parameter is deprecated"
-    ):
-        f(1)
-
 
 def test_no_validation():
     """Check that validation can be skipped for a parameter."""