[ENH] Added possibility for pooling strides in TimeCNN (#2485)

* Added strides_pooling in CNNNetwork

* Added strides_pooling in CNNClassifier

* Added strides_pooling in docstrings

* Automatic `pre-commit` fixes

* Line Limit

---------

Co-authored-by: kavya-r30 <kavya-r30@users.noreply.github.com>

aeon/classification/deep_learning/_cnn.py

@@ -42,6 +42,9 @@ class TimeCNNClassifier(BaseDeepClassifier):
     strides : int or list of int, default = 1
         The strides of kernels in the convolution and max pooling layers, if not a
         list, the same strides are used for all layers.
+    strides_pooling : int or list of int, default = None
+        Strides for the pooling layers. If None, defaults to pool_size.
+        If not a list, the same strides are used for all pooling layers.
     dilation_rate : int or list of int, default = 1
         The dilation rate of the convolution layers, if not a list, the same dilation
         rate is used all over the network.
@@ -125,6 +128,7 @@ def __init__(
         activation="sigmoid",
         padding="valid",
         strides=1,
+        strides_pooling=None,
         dilation_rate=1,
         n_epochs=2000,
         batch_size=16,
@@ -148,6 +152,7 @@ def __init__(
         self.n_filters = n_filters
         self.padding = padding
         self.strides = strides
+        self.strides_pooling = strides_pooling
         self.dilation_rate = dilation_rate
         self.avg_pool_size = avg_pool_size
         self.activation = activation
@@ -182,6 +187,7 @@ def __init__(
             activation=self.activation,
             padding=self.padding,
             strides=self.strides,
+            strides_pooling=self.strides_pooling,
             dilation_rate=self.dilation_rate,
             use_bias=self.use_bias,
         )

aeon/networks/_cnn.py

@@ -32,6 +32,9 @@ class TimeCNNNetwork(BaseDeepLearningNetwork):
     strides : int or list of int, default = 1
         The strides of kernels in the convolution and max pooling layers, if not a list,
         the same strides are used for all layers.
+    strides_pooling : int or list of int, default = None
+        Strides for the pooling layers. If None, defaults to pool_size.
+        If not a list, the same strides are used for all pooling layers.
     dilation_rate : int or list of int, default = 1
         The dilation rate of the convolution layers, if not a list, the same dilation
         rate is used all over the network.
@@ -65,6 +68,7 @@ def __init__(
         activation="sigmoid",
         padding="valid",
         strides=1,
+        strides_pooling=None,
         dilation_rate=1,
         use_bias=True,
     ):
@@ -75,6 +79,7 @@ def __init__(
         self.activation = activation
         self.padding = padding
         self.strides = strides
+        self.strides_pooling = strides_pooling
         self.dilation_rate = dilation_rate
         self.use_bias = use_bias
 
@@ -131,6 +136,19 @@ def build_network(self, input_shape, **kwargs):
         else:
             self._avg_pool_size = [self.avg_pool_size] * self.n_layers
 
+        if self.strides_pooling is None:
+            self._strides_pooling = self._avg_pool_size
+        elif isinstance(self.strides_pooling, list):
+            if len(self.strides_pooling) != self.n_layers:
+                raise ValueError(
+                    f"Number of strides for pooling {len(self.strides_pooling)}"
+                    f" should be the same as number of layers but is"
+                    f" not: {self.n_layers}"
+                )
+            self._strides_pooling = self.strides_pooling
+        else:
+            self._strides_pooling = [self.strides_pooling] * self.n_layers
+
         if isinstance(self.activation, list):
             if len(self.activation) != self.n_layers:
                 raise ValueError(
@@ -204,9 +222,10 @@ def build_network(self, input_shape, **kwargs):
                 use_bias=self._use_bias[i],
             )(x)
 
-            conv = tf.keras.layers.AveragePooling1D(pool_size=self._avg_pool_size[i])(
-                conv
-            )
+            conv = tf.keras.layers.AveragePooling1D(
+                pool_size=self._avg_pool_size[i],
+                strides=self._strides_pooling[i],
+            )(conv)
 
             x = conv