Diabetes Predictor Using Machine Learning

DISCLAIMER: This program is made for demonstration and
entertainment purposes, it is NOT a real medical diabetes predictor.

Preview Video:

I have implemented a small diabetes predictor from data of more than 768 female patients, I have taken the data from the web:

https://www.kaggle.com/uciml/pima-indians-diabetes-database/data
It is a database of diabetes patients from India.

The algorithm that I have implemented uses Machine Learning, specifically the K-Neighbors technique.

At the time of this writing, there is a data set of 768 women patients, this is the summary of the data:

	Pregnancies	Glucose	BloodPressure	SkinThickness	Insulin	BMI	DiabetesPedigreeFunction	Age	Outcome
count	768.000000	768.000000	768.000000	768.000000	768.000000	768.000000	768.000000	768.000000	768.000000
mean	3.845052	120.894531	69.105469	20.536458	79.799479	31.992578	0.471876	33.240885	0.348958
std	3.369578	31.972618	19.355807	15.952218	115.244002	7.884160	0.331329	11.760232	0.476951
min	0.000000	0.000000	0.000000	0.000000	0.000000	0.000000	0.078000	21.000000	0.000000
max	17.000000	199.000000	122.000000	99.000000	846.000000	67.100000	2.420000	81.000000	1.000000

• Accuracy: 0.7083333333333334


• Train size: 576


• Test size: 192

Instructions

It works as follows, I have made a web form in which you have to fill out the following data:

• Pregnancies: Number of times a patient has become pregnant, in the database there are female patients who have become pregnant up to 17 times (it does not mean that they have had 17 children).


• Glucose: Blood glucose level, there are some patients with levels up to 199 mg/dL (milligrams per deciliter).


• BloodPressure: Blood pressure, measured in mm Hg, there are some patients with levels up to 122 mm Hg.


• SkinThickness: Skin thickness (fat layer under the skin), it is measured using ultrasound or calipers, there are some patients with levels up to 99 mm.


• Insulin: Insulin level in blood, normal levels are below 100 mg/dL (milligrams per deciliter), above the 126 mg/dL is considered diabetes, there are some patients with levels up to 846 mg/dL.


• BMI: Body mass index, also called body mass index, Normal levels are between 18.5 and 25 kg/m², there are some patients with levels up to 67.1 kg/m².


• DiabetesPedigreeFunction: it is a function that measures the probability of having diabetes based on family history. This function receives multiple parameters and returns zero if you have no probability or one if you have it. There are some patients with levels up 2.42.


• Age: The age of the patient, including newborns (0 years old) and elderly (81 years or older).

Once the form is filled in, press Submit and the machine learning algorithm will calculate a prediction, in this way it will indicate if the patient is prone to having diabetes or not.

You can see that in the following image it appears that the patient is not prone to diabetes.

DISCLAIMER: This program is made for demonstration and
entertainment purposes, it is NOT a real medical diabetes predictor.

See you in the next post!

References:

• https://www.kaggle.com/uciml/pima-indians-diabetes-database/data
• Python Machine Learning for Beginners
  BN ID: 2940161229385
  Publisher: AI SCIENCES LLC
  Publication date: 01/05/2019
  

Jorge Gascon Perez

Back