Diabetes classification with deployment

• Diabetes classification using the Prima Indians Diabetes Database
• Dataset: https://www.kaggle.com/datasets/uciml/pima-indians-diabetes-database
• Tried various classification algorithms.
• Imporved the accuracy of the models by:
  • Removing outliers from the data
  • Adjusting the hyperparametes of the models
• Currently, the best model is logistic regression with accuracy: 82.21%, Precision: 77.50%, Recall: 60.78%, and F1-Score: 0.68.

About dataset:

Context

This dataset is originally from the National Institute of Diabetes and Digestive and Kidney Diseases. The objective of the dataset is to diagnostically predict whether or not a patient has diabetes, based on certain diagnostic measurements included in the dataset. Several constraints were placed on the selection of these instances from a larger database. In particular, all patients here are females at least 21 years old of Pima Indian heritage.

Content

The datasets consists of several medical predictor variables and one target variable, Outcome. Predictor variables includes the number of pregnancies the patient has had, their BMI, insulin level, age, and so on.

Tech stack used

1. Python
2. Flask
3. HTML
4. AWS

Deployment link

http://diabetes-env.eba-mngrqrji.eu-north-1.elasticbeanstalk.com/predictdata

Screenshot of the webapp

To run offline:

This can also be run offline. To do so, follow the steps below:

Clone the repo

git clone 'https://github.com/LoyumM/Diabetes-classification-with-cloud-deployment.git'

Create a virtual environment

conda create -p venv python == 3.8 -y

Activate environment

conda activate venv/

Install the required modules

pip install -r requirements.txt

Run the application

python application.py

Open the following link in a browser: http://127.0.0.1:5000/predictdata