Stroke Prediction

Inspiration

The inspiration of this project was derived from the theme of the datathon : Tech for Good

What it does

The project predicts if a respective individual will encounter a stroke or not based on their lifestyle. The dataset consists of 11 clinical attributes that define everyday lifestyle of a person and based on those given features we can predict their stroke likelihood.

How we built it

The model was built using the TensorFlow library in Python. We used a feed-forward neural network to classify the inputs.

There are a total of 6 hidden layers in the model with three dropout layers to avoid overfitting of the data.

The final accuracy as such of the model came out to be around 72%

What's next for Stroke Prediction

We want to further develop this model with a front-end to make it available to the general public as well as simultaneously improve the accuracy of the model to get better predictions out of it.

Built With

python
tensorflow

Submitted to

IBM zDatathon

Created by

I worked on the TensorFlow model

Tanay Singh
I worked on the front-end part which is a web app which takes inputs of the data prediction model and the ML processes it and then the final prediction is shown in the output page of the front-end

Raghav Loknath
I worked on the creation of the ml model and integration of front end with backend using Flask and Pickle module

Sudhir Manivannan
I worked on ML model deployment and flask integration

vishnudeep14
I worked on the Front-end part of the project which is basically the starting page of dis project using HTML nd CSS

Vishwas M
I was one of the people who worked on the web page/web application. I coded the html pages and their css parts. This webpage forms the essential link between the predicting AI ML model and the user.

Vishal CE

Updates

Tanay Singh started this project — Sep 25, 2022 02:40 AM EDT

Leave feedback in the comments!

Log in or sign up for Devpost to join the conversation.