top of page

A sigmoidal regression for determining high risk patients

 A sigmoidal regression is a way of determining if a single observation stands as a failure or a succes in a probability analysis, for determining if a patient might be of high risk one or a normal one this kind of tools could be use. In this case a "success" is when a patient is a high risk one, and a failure when a patient is not.

sig.PNG

Multiple Linear Regression

Multi.PNG

For determining the result of the sigmoidal, a buch of variables must be cinsider such as: weight, age, gender, if the patient has diabetes or another kind of comorbility. (For determining the most important comorbilities we could use a corraltion matrix). 

​

After the multiple linear regression(MLR) is done we could mix the idea of the sigmoidal function and the MLR, in order to determine if a pacient with a specific sets of characteritics could be classified as a high risk patient or not.

E.PNG

What is next?

Find some repository such as Kaggle or GitHub in order to download real data and perform the sigmoidal analysis.

+ 52 55 49 55 31 37

  • linkedin

©2020 por Data Science: Models and Applications Creada con Wix.com

bottom of page