update

jizhang02 · Sep 3, 2024 · 9751d0c · 9751d0c
1 parent 39acea8
commit 9751d0c
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 2 deletions.
diff --git a/deep_learning_from_scratch.ipynb b/deep_learning_from_scratch.ipynb
@@ -6434,7 +6434,11 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "### Pooling layer"
+    "### Pooling layer\n",
+    "The pooling layer is used to reduce the size of the feature map, which can reduce the number of parameters and the amount of computation. The pooling layer is different from the convolutional layer in that it does not have parameters to learn. Pooling simply takes the maximum value (or average) from the target region, so there are no parameters to learn. The number of channels remains the same. Robust to small positional changes. For example, the figure below.    \n",
+    "![pooling](./figures/dlscratch_poolposition.png)    \n",
+    "Max pooling    \n",
+    "![max pooling](./figures/dlscratch_poolmax.png)    "
    ]
   },
   {
@@ -6496,7 +6500,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.6"
+   "version": "3.11.5"
   },
   "vscode": {
    "interpreter": {

diff --git a/figures/dlscratch_poolmax.png b/figures/dlscratch_poolmax.png
diff --git a/figures/dlscratch_poolposition.png b/figures/dlscratch_poolposition.png