Predicting Strokes

Abstract (Draft)

According to the CDC, in 2020, someone in the United States has a stroke every 40 seconds. Every 4 minutes, someone dies of a stroke. I plan to use the Kaggle Stroke Prediction Dataset. Research has shown that if a stroke-patient went into therapy in the first few weeks, almost complete recovery is possible (Krakauer, 2015). That is because right after a stroke, the brain’s plasticity increases–which is the brain’s ability to-reform re-shape, and re-organize. So any damaged part can be compensated for.

Additionally, not only are strokes so damaging, but silent strokes are strokes that have no appearnt symptoms and can cause permanent brain damage without the person realizing. I wanted to look at whether age and gender be enough to predict the likely-hood of developing a stroke?

https://drive.google.com/file/d/18cWfaBepn1zLWRWWo0TCYWB2mtWw_p7t/view?usp=sharing

P.S. I am so sorry if this was boring– I wasn’t feeling well. Also, I accidentally say at the end that my model is very good; I meant in terms of accuracy and loss values. However, I don’t think it’s even close to accurately predicting actual stroke cases, especially that it mainly relied on age. It is also obvious that it has some issues, as I noted in the video.