batch norm

content/wiki/machine-learning/neural-networks/batch-norm/index.md

@@ -0,0 +1,26 @@
+---
+title: "Batch Norm"
+description: "Batch norm"
+date: 2021-02-15
+categories:
+  - "Machine Learning"
+tags:
+  - "Machine Learning"
+references:
+  - name: "
+    1.Rohrer B. Batch normalization [Internet]. E2EML School. [cited 2023 Oct 22]. Available from: https://e2eml.school/batch_normalization.html"
+    link: "https://e2eml.school/batch_normalization.html"
+    key: "Rohrer-Batch-Normalization"
+links:
+  - wiki/model-selection/mdl-and-neural-networks.md
+weight: 6
+---
+
+
+Batch norm is a normalization method to relieve the internal covariate shift[^Rohrer-Batch-Normalization].
+
+
+[David Page created a colab notebook to](https://colab.research.google.com/github/davidcpage/cifar10-fast/blob/master/batch_norm_post.ipynb#scrollTo=yXu58M6ZLayL) provide some real examples of how batch norm helps with the internal covariance shift.
+
+
+[^Rohrer-Batch-Normalization]: {{< cite key="Rohrer-Batch-Normalization" >}}