AWS prescriptive-guidance documentation change
Summary
Updated Apache Spark documentation link version from 3.5.3 to latest
Security assessment
Change only updates a version number in a documentation link for Apache Spark. No security implications or content changes.
Diff
diff --git a/prescriptive-guidance/latest/ml-operations-planning/splits-leakage.md b/prescriptive-guidance/latest/ml-operations-planning/splits-leakage.md index f60ee28ec..401e08adb 100644 --- a//prescriptive-guidance/latest/ml-operations-planning/splits-leakage.md +++ b//prescriptive-guidance/latest/ml-operations-planning/splits-leakage.md @@ -21 +21 @@ By splitting your data in this way, you can use the `validation` set to choose a -When you split your data into `training`, `validation`, and `testing` for small datasets, or when you work with highly imbalanced data, make sure to use a stratified split algorithm. Stratification guarantees that each split contains approximately the same number or distribution of classes for each split. The [scikit-learn ML library](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html) already implements stratification, and so does [Apache Spark](https://spark.apache.org/docs/3.5.3/api/python/reference/pyspark.sql/api/pyspark.sql.DataFrame.sampleBy.html). +When you split your data into `training`, `validation`, and `testing` for small datasets, or when you work with highly imbalanced data, make sure to use a stratified split algorithm. Stratification guarantees that each split contains approximately the same number or distribution of classes for each split. The [scikit-learn ML library](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html) already implements stratification, and so does [Apache Spark](https://spark.apache.org/docs/latest/api/python/reference/pyspark.sql/api/pyspark.sql.DataFrame.sampleBy.html).