diff --git a/python/docs/pyspark.ml.rst b/python/docs/pyspark.ml.rst
index 9015a3c15edb2..f20453e624296 100644
--- a/python/docs/pyspark.ml.rst
+++ b/python/docs/pyspark.ml.rst
@@ -5,9 +5,33 @@ Submodules
 ----------
 
 pyspark.ml module
--------------------------
+-----------------
 
 .. automodule:: pyspark.ml
     :members:
     :undoc-members:
     :show-inheritance:
+
+pyspark.ml.param module
+-----------------------
+
+.. automodule:: pyspark.ml.param
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+pyspark.ml.feature module
+-------------------------
+
+.. automodule:: pyspark.ml.feature
+    :members:
+    :undoc-members:
+    :show-inheritance:
+
+pyspark.ml.classification module
+--------------------------------
+
+.. automodule:: pyspark.ml.classification
+    :members:
+    :undoc-members:
+    :show-inheritance:
diff --git a/python/pyspark/ml/__init__.py b/python/pyspark/ml/__init__.py
index df39515d81433..83351dc523ce1 100644
--- a/python/pyspark/ml/__init__.py
+++ b/python/pyspark/ml/__init__.py
@@ -22,7 +22,7 @@
 from pyspark.ml.param import Param, Params
 from pyspark.ml.util import Identifiable
 
-__all__ = ["Pipeline", "Transformer", "Estimator"]
+__all__ = ["Pipeline", "Transformer", "Estimator", "param", "feature", "classification"]
 
 
 def _jvm():
diff --git a/python/pyspark/ml/classification.py b/python/pyspark/ml/classification.py
index 2c9aaad03cedf..f3cea0958c897 100644
--- a/python/pyspark/ml/classification.py
+++ b/python/pyspark/ml/classification.py
@@ -15,11 +15,12 @@
 # limitations under the License.
 #
 
-from pyspark.sql import SchemaRDD
+from pyspark.sql import SchemaRDD, inherit_doc
 from pyspark.ml import Estimator, Transformer, _jvm
 from pyspark.ml.param import Param
 
 
+@inherit_doc
 class LogisticRegression(Estimator):
     """
     Logistic regression.
diff --git a/python/pyspark/ml/feature.py b/python/pyspark/ml/feature.py
index 35f32f513f7a5..9e4b6574574a4 100644
--- a/python/pyspark/ml/feature.py
+++ b/python/pyspark/ml/feature.py
@@ -15,10 +15,11 @@
 # limitations under the License.
 #
 
-from pyspark.sql import SchemaRDD, ArrayType, StringType
+from pyspark.sql import SchemaRDD, ArrayType, StringType, inherit_doc
 from pyspark.ml import Transformer, _jvm
 from pyspark.ml.param import Param
 
+@inherit_doc
 class Tokenizer(Transformer):
 
     def __init__(self):
@@ -61,6 +62,7 @@ def transform(self, dataset, params={}):
             raise ValueError("The input params must be either a dict or a list.")
 
 
+@inherit_doc
 class HashingTF(Transformer):
 
     def __init__(self):