This is a colaboratory notebook. Stroke-Prediction uses the healthcare-dataset-stroke-data dataset to predict the probability of stroke in a patient based on other factors such as age, gender, work type, residence type and illness history.
Using pandas, some conclusions were made about the effects of these factors on the likelihood of having stroke.
Then a predictive model was built using TensorFlow.
The data was first split into a training set and a testing set. Then we were able to train our model using our training data, and then test it using our testing data.
The model had an accuracy rate of 94%