Upgrade XGBoost to >=1.4.0

docs/source/release_notes.rst

@@ -9,6 +9,7 @@ Release Notes
         * Fixed ``FraudCost`` objective and reverted threshold optimization method for binary classification to ``Golden`` :pr:`2450`
         * Added custom exception message for partial dependence on features with scales that are too small :pr:`2455`
     * Changes
+        * Updated XGBoost dependency to ``>=1.4.0`` :pr:`2484`
     * Documentation Changes
         * Moved docstrings from ``__init__`` to class pages, added missing docstrings for missing classes, and updated missing default values :pr:`2452`
     * Testing Changes

evalml/tests/dependency_update_check/latest_dependency_versions.txt

@@ -29,4 +29,4 @@ sktime==0.6.1
 statsmodels==0.12.2
 texttable==1.6.3
 woodwork==0.4.2
-xgboost==1.2.1
+xgboost==1.4.2

requirements.txt

@@ -2,7 +2,7 @@
 plotly>=5.0.0
 kaleido>=0.1.0
 ipywidgets>=7.5
-xgboost>=1.1.0,<1.3.0
+xgboost>=1.4.0
 catboost>=0.20
 lightgbm>=2.3.1
 matplotlib>=3.3.3