From 39c2433a0161341b7ad8e084320d994708ffd78f Mon Sep 17 00:00:00 2001 From: Paulo Cilas Morais Lyra Junior <58008200+paulocilasjr@users.noreply.github.com> Date: Wed, 11 Dec 2024 11:36:39 -0500 Subject: [PATCH] Update topics/statistics/tutorials/loris_model/tutorial.md Co-authored-by: Anup Kumar, PhD --- topics/statistics/tutorials/loris_model/tutorial.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/topics/statistics/tutorials/loris_model/tutorial.md b/topics/statistics/tutorials/loris_model/tutorial.md index 88dd6b47dd106f..1492d490449ad4 100644 --- a/topics/statistics/tutorials/loris_model/tutorial.md +++ b/topics/statistics/tutorials/loris_model/tutorial.md @@ -86,7 +86,7 @@ The neutrophil–lymphocyte ratio (NLR), a biomarker derived from the ratio of n In this dataset, NLR values range, for example, from 0.8 to 88, with several extreme outliers. To address this, NLR values will be truncated at 25, meaning any value above 25 will be capped at 25. This truncation is important for preventing extreme outliers from disproportionately influencing the machine learning model. ## Age -n predictive models for patient outcomes, age is a crucial feature because it is often correlated with various health factors and disease risks. As people age, their immune systems, metabolism, and ability to recover from illnesses may change, influencing how they respond to treatments, medications, or disease progression. Including age as a feature helps models account for the biological changes that occur over time and can improve the accuracy of predictions across different age groups. +In predictive models for patient outcomes, age is a crucial feature because it is often correlated with various health factors and disease risks. As people age, their immune systems, metabolism, and ability to recover from illnesses may change, influencing how they respond to treatments, medications, or disease progression. Including age as a feature helps models account for the biological changes that occur over time and can improve the accuracy of predictions across different age groups. However, there are limits to how predictive age might be, particularly for extreme values. For example, patients over a certain age may share similar health characteristics, and further increases in age may not significantly add predictive value. Truncating age to a maximum value (like 85) helps avoid overemphasizing small differences between very old patients, where the added predictive power might be negligible.