[DOC] Compose and deep learning classifier doc tidy (#3756)

* compose package comments

* compose package comments

* deep learning package comments

sktime/classification/compose/_column_ensemble.py

@@ -189,35 +189,30 @@ class ColumnEnsembleClassifier(BaseColumnEnsembleClassifier):
 
     This estimator allows different columns or column subsets of the input
     to be transformed separately and the features generated by each
-    transformer
-    will be ensembled to form a single output.
+    transformer will be ensembled to form a single output.
 
     Parameters
     ----------
     estimators : list of tuples
-        List of (name, estimator, column(s)) tuples specifying the
-        transformer objects to be applied to subsets of the data.
+        List of (name, estimator, column(s)) tuples specifying the transformer
+        objects to be applied to subsets of the data.
 
         name : string
             Like in Pipeline and FeatureUnion, this allows the
-            transformer and
-            its parameters to be set using ``set_params`` and searched
-            in grid
-            search.
+            transformer and its parameters to be set using ``set_params`` and searched
+            in grid search.
         estimator :  or {'drop'}
             Estimator must support `fit` and `predict_proba`. Special-cased
             strings 'drop' and 'passthrough' are accepted as well, to
-            indicate to drop the columns
-        column(s) : string or int, array-like of string or int, slice, \
-            boolean mask array or callable
+            indicate to drop the columns.
+        column(s) : array-like of string or int, slice, boolean mask array or callable.
 
     remainder : {'drop', 'passthrough'} or estimator, default 'drop'
         By default, only the specified columns in `transformations` are
         transformed and combined in the output, and the non-specified
         columns are dropped. (default of ``'drop'``).
         By specifying ``remainder='passthrough'``, all remaining columns
-        that
-        were not specified in `transformations` will be automatically passed
+        that were not specified in `transformations` will be automatically passed
         through. This subset of columns is concatenated with the output of
         the transformations.
         By setting ``remainder`` to be an estimator, the remaining

sktime/classification/compose/_ensemble.py

@@ -28,24 +28,10 @@
 
 
 class ComposableTimeSeriesForestClassifier(BaseTimeSeriesForest, BaseClassifier):
-    """Time-Series Forest Classifier.
-
-    @article{DENG2013142,
-        title = {A time series forest for classification and feature extraction},
-        journal = {Information Sciences},
-        volume = {239},
-        pages = {142 - 153},
-        year = {2013},
-        issn = {0020-0255},
-        doi = {https://doi.org/10.1016/j.ins.2013.02.030},
-        url = {http://www.sciencedirect.com/science/article/pii/S0020025513001473},
-        author = {Houtao Deng and George Runger and Eugene Tuv and Martyanov Vladimir},
-        keywords = {Decision tree, Ensemble, Entrance gain, Interpretability,
-                    Large margin, Time series classification}
-    }
+    """Time Series Forest Classifier as described in [1]_.
 
-    A time series forest is a meta estimator and an adaptation of the random
-    forest for time-series/panel data that fits a number of decision tree
+    A time series forest is an adaptation of the random
+    forest for time-series data. It that fits a number of decision tree
     classifiers on various sub-samples of a transformed dataset and uses
     averaging to improve the predictive accuracy and control over-fitting.
     The sub-sample size is always the same as the original input sample size
@@ -182,6 +168,11 @@ class labels (multi-output problem).
         set. If n_estimators is small it might be possible that a data point
         was never left out during the bootstrap. In this case,
         `oob_decision_function_` might contain NaN.
+
+    References
+    ----------
+    .. [1] Deng et. al, A time series forest for classification and feature extraction,
+    Information Sciences, 239:2013.
     """
 
     _tags = {

sktime/classification/deep_learning/cnn.py

@@ -14,7 +14,7 @@
 
 
 class CNNClassifier(BaseDeepClassifier):
-    """Time Convolutional Neural Network (CNN), as described in [1].
+    """Time Convolutional Neural Network (CNN), as described in [1]_.
 
     Parameters
     ----------
@@ -49,13 +49,14 @@ class CNNClassifier(BaseDeepClassifier):
 
     Notes
     -----
-    .. [1] Zhao et. al, Convolutional neural networks for
-    time series classification, Journal of
-    Systems Engineering and Electronics, 28(1):2017.
-
     Adapted from the implementation from Fawaz et. al
     https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/cnn.py
 
+    References
+    ----------
+    .. [1] Zhao et. al, Convolutional neural networks for time series classification,
+    Journal of Systems Engineering and Electronics, 28(1):2017.
+
     Examples
     --------
     >>> from sktime.classification.deep_learning.cnn import CNNClassifier

sktime/classification/deep_learning/fcn.py

@@ -14,7 +14,7 @@
 
 
 class FCNClassifier(BaseDeepClassifier):
-    """Fully Connected Neural Network (FCN), as described in [1].
+    """Fully Connected Neural Network (FCN), as described in [1]_.
 
     Parameters
     ----------
@@ -42,12 +42,13 @@ class FCNClassifier(BaseDeepClassifier):
 
     Notes
     -----
-    .. [1] Zhao et. al, Convolutional neural networks for
-    time series classification, Journal of
-    Systems Engineering and Electronics, 28(1):2017.
-
     Adapted from the implementation from Fawaz et. al
-    https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/cnn.py
+    https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/fcn.py
+
+    References
+    ----------
+    .. [1] Zhao et. al, Convolutional neural networks for time series classification,
+    Journal of Systems Engineering and Electronics, 28(1):2017.
 
     Examples
     --------

sktime/classification/deep_learning/mlp.py

@@ -14,7 +14,7 @@
 
 
 class MLPClassifier(BaseDeepClassifier):
-    """Multi Layer Perceptron Network (MLP), as described in [1].
+    """Multi Layer Perceptron Network (MLP), as described in [1]_.
 
     Parameters
     ----------
@@ -42,16 +42,15 @@ class MLPClassifier(BaseDeepClassifier):
 
     Notes
     -----
-    .. .. [1]  Network originally defined in:
-    @inproceedings{wang2017time, title={Time series classification from
-    scratch with deep neural networks: A strong baseline}, author={Wang,
-    Zhiguang and Yan, Weizhong and Oates, Tim}, booktitle={2017
-    International joint conference on neural networks (IJCNN)}, pages={
-    1578--1585}, year={2017}, organization={IEEE} }
-
     Adapted from the implementation from source code
     https://github.com/hfawaz/dl-4-tsc/blob/master/classifiers/mlp.py
 
+    References
+    ----------
+    .. [1] Wang et. al, Time series classification from
+    scratch with deep neural networks: A strong baseline,
+    International joint conference on neural networks (IJCNN), 2017.
+
     Examples
     --------
     >>> from sktime.classification.deep_learning.mlp import MLPClassifier

sktime/classification/deep_learning/tapnet.py

@@ -20,7 +20,7 @@
 
 
 class TapNetClassifier(BaseDeepClassifier):
-    """Implementation of TapNetClassifier, as described in [1].
+    """Time series attentional prototype network (TapNet), as described in [1]_.
 
     Parameters
     ----------