What's Stroking Your Brain?


A Supervised Machine Learning Analysis

The Premise

Millennials are getting old.

That's right.

And with great age, comes a never-ending array of probable health issues, with one of those health issues being strokes.

Trying to see if the triggers of a stroke can be identified, a supervised machine learning model was created and implemented on a stroke prediction dataset from kaggle.

Once the model was created, the Top 3 variables (which made up 77% of importance determined by our model *out of a total of 13*) were chosen as the basis of the website, which was designed to be nostalgia-inducing to Millennials. It contains a quiz, interactive visuals, and social commentary while being written in a friendly, casual tone to appeal to this generation that is now entering their 40's. The name and logo of the site is even a play on a very popular website that is still around today.

Click Here To Visit The fUZZbEED Page!

Click Here To Visit The GitHub Page!

GitHub

Languages and Programs Used:

  • Python

    • Language used in the notebooks

  • HTML

    • Language used to build the index.html

  • CSS

    • Language used to style the website

  • Javascript

    • Language used to program the interactive components of the page

  • Google Colaboratory

    • Notebook used to create and run the machine learning model

  • Jupyter Notebook

    • Notebook used to create the analysis page

  • Tableau

    • Program used to create the visuals

  • PANDAS

    • Python library used to clean up the data and to assign numerical values to our string/object data

  • Sklearn

    • Python library used to create the Random Forest Classifier model

    Screenshots:

    Dashboard

    Dashboard

    Dashboard