DiaTect
Check your diabetes status by answering 16 qualifying questions.
Embedded Files
Careful, this website is still under construction.
Overview
Overview
DiaTect is a Neural Network-based model trained based on the age, gender and 14 symptoms of 540 diabetic and non-diabetic patients. It achieved an accuracy of 97.67% and an f1-score of 0.98. Streamlit-deployed model can be accessed at the bottom of the page.
Dataset features
Dataset features
Complexity
16 training features with binary target class
Diversity
Patients of varying ages from 16 to 90
Volume
Total number of training examples: 540
Format
csv
Demo
Raw data
Project details
Project details
Type: Side-project
Status: Completed
Technologies used: Python | Streamlit | Scikit-learn
Tasks:
Tasks:
Binary encoding
Binary classification
Model pickling
Deployment with streamlit
Try it out:
Try it out:
Please wait a few seconds to try the model below. Turn device to landscape mode if you are on mobile for ease of use.
Page updated
Google Sites
Report abuse