According to the Center for Disease and Prevention (CDC) Every 40 seconds, someone in the United States has a stroke. Every 3.5 minutes, someone dies of stroke. This project uses parameters like gender, age, various diseases, and smoking status to predict if a patient had a stroke. Each row in the data provides relavant information about the patient.
To discover the drivers of stroke to help our patients prevent a stroke occurance.
- 4.86% of the patient population has suffered a stroke
- Patients wih heart disease have a larger increase in stroke rate than patients with hypertension
- The patients age is a driver of stroke
- Help patients control medical conditions to reduce the chances of stroke
- Give patients the tools and guidance to treat heart conditions
- Encourage patients to keep track of their blood pressure and take appropriate medication when needed
In order to improve model more data collection and health information such as:
- Patients family history of stroke or heart disease
- Demographics of the patients
Discover drivers of stroke. Use drivers to develop a machine learning model to predict weather or not a patient had a stroke. Use findings to see what preventative measures if any can be taken to prevent a stroke.
My initial hypothesis is that hypertension is a driver of stroke.
-
Acquire data from Kaggle
-
Prepare data
-
Explore data in search of drivers of stroke
-
Answer the following initial questions:
-
- What is the percent of patients who have suffered a stroke?
-
- Does the presense of hypertension increase the risk of stroke?
-
- Are patients with a heart condition more at risk of stroke than patients with hypertension?
-
- Controling for gender of a patient, does heart disease increases risk of stroke?
-
- Is age a driver of stroke?
-
- Do patients who have ever been married suffer more strokes than patients that have not been married?
-
-
Use drivers identified in explore to build predictive models of different types
- Evaluate models on train and validate data
- Select the best model based on highest accuracy
- Evaluate the best model on test data
-
Draw conclusions
| Target Variable | Description |
|---|---|
| stroke | If the patient has a stroke or not. (1 = yes, 0 = no) |
| Feature | Description |
|---|---|
| id | unique identifier |
| gender | Gender of patient Male or Female or Other |
| age | age of the patient |
| hypertension | If a patient has hypertension (1 = yes, 0 = no) |
| heart_disease | If a patient has any heart diseases (1 = yes, 0 = no) |
| ever_married | If a patient has ever been married (1 = yes, 0 = no) |
| Additional Features | Encoded and values for categorical data and scaled versions continuous data |
1. Clone this repository
2. Get Telco Churn data in a csv from Kaggle:
https://www.kaggle.com/datasets/fedesoriano/stroke-prediction-dataset?select=healthcare-dataset-stroke-data.csv
3. Save file in cloned repository
4. Run notebook
- 4.86% of the patient population has suffered a stroke
- Patients with hypertension have an increase in stroke rate
- Patients with heart disease have an increase in stroke rate
- Patients with heart disease have a larger increase in stroke rate than patients with hypertension
- The male gender of our patients that have heart disease have a higher stroke rate than the female gender
- The patients age is a driver of stroke
- Patients who have been married have a slight increase in stroke rate.
In order to improve model more data collection and health information such as:
- if the patients family has a history of stroke or heart disease
- Demographics of the patients
- Patients weight and height
Current top model is not ready to perform.